Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Sulphur management strategies for anaerobic treatment of a mechanical pulping effluent Stephenson, Robert John 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1994-893869.pdf [ 5MB ]
Metadata
JSON: 831-1.0058546.json
JSON-LD: 831-1.0058546-ld.json
RDF/XML (Pretty): 831-1.0058546-rdf.xml
RDF/JSON: 831-1.0058546-rdf.json
Turtle: 831-1.0058546-turtle.txt
N-Triples: 831-1.0058546-rdf-ntriples.txt
Original Record: 831-1.0058546-source.json
Full Text
831-1.0058546-fulltext.txt
Citation
831-1.0058546.ris

Full Text

SULPHUR MANAGEMENT STRATEGIESFOR ANAEROBIC TREATMENTOF A MECHAMCAL PULPING EFFLUENTbyROBERT JOHN STEPHENSONB.A.Sc., Chemical Engineering, Queen’s University at Kingston, 1980M.A.Sc., Bio-Resource Engineering, University of British Columbia, 1987A THESIS SUBMITTED iN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUD]ESDepartment of Chemical EngineeringWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIANovember 1993© Robert John Stephenson, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(SignatureDepartment of_______________________c-)The University of British ColumbiaVancouver, CanadaDate_____DE-6 (2/88)AbstractPulp manufacture uses sulphur in a variety of forms and some form of thesesulphur compounds ultimately appears in the effluent. Under anaerobic conditions,sulphate, sulphite and thiosuiphate are reduced to sulphide. This results in problems oftoxicity, odour, corrosion, and inhibition of the wastewater treatment microorganisms.The fate of these inorganic sulphur compounds in a bleached chemithermomechanicalpuip/thermomechanical pulp (BCTMPITMP) effluent mixture was examined in two phaseanaerobic reactors at 35 °C and 55 °C. The following sulphur management strategies wereinvestigated: 1) shifting the sulphide to the less toxic form by controlling the pH of theacidogenic reactor, 2) inhibiting the sulphur reducing bacteria via molybdenum addition tothe feed tank, and 3) stripping the hydrogen sulphide dissolved in the methane phasereactor liquor by recycling hydrogen sulphide-free scrubbed off gas. The laboratory scaleexperimental apparatus consisted of upflow anaerobic sludge bed pre-treatment oracidogenic reactors followed by hybrid upflow anaerobic sludge bed/fixed filmmethanogenic reactors.The sulphur management strategies which were investigated demonstratedsignificantly improved treatment efficiencies.At 35 °C, controlling the pH of the acidogenic reactors with sodium carbonatefrom 5.5 (uncontrolled) to 8.0 in order to shift the formed sulphide species to the lesstoxic ionic form appeared to be effective in promoting the wastewater treatmentefficiency. Sulphate reduction efficiencies were typically 75 to 90% irrespective of acidphase reactor pH. Maximum total organic carbon (TOC) removals of 55 and 63% wereobserved at an acid phase reactor pH of 7.5 for total hydraulic retention times (HRTs) of1.2 and 1.8 days respectively.Molybdate was added to the wastewater at levels from 0.1 to 1.0 mM. At 35 °C, itwas effective in two out of three effluent batches at the 1.0 mM level in decreasing11sulphate reduction from 90% down to 23 to 40%. Maximum TOC removal efficiencies of42% were observed at 0.5 mM molybdate. Molybdate additions greater than 0.5 mMresulted in reduced TOC removals and gas production rates.Hydrogen suiphide stripping, using ferric chloride scrubbed and recycled off gas,resulted in lower dissolved suiphide levels and increased TOC removals. Sulphatereduction was unaffected by the varying concentrations of dissolved suiphide. Strippingresulted in TOC removal efficiencies of up to 57%, a significant improvement over theunstripped control runs where the TOC removals were only approximately 24%. Thissulphur management strategy at 35 °C appeared to be the most effective means of sulphurmanagement for sulphur rich mechanical pulping effluents.Thermophilic 55 °C anaerobic treatment was also studied using the same effluent,inocula and sulphur management strategies. Overall, both the treatment efficiency and thesulphate reduction were considerably lower for the thermophilic runs compared to themesophilic runs. Raising the acidogenic phase reactor pH from 7.0 to 7.5 to 8.0 appearedto have no significant effect on the organic carbon removal efficiency (maximum 24%) oron sulphate reduction efficiency where a maximum of only 51% was realized.Thermophilic molybdate inhibition of sulphate reduction was not as marked as for the 1.0mM level at 35 °C, perhaps due to the already low baseline sulphate reduction efficiency(maximum of 50%) at 55 °C. Molybdate addition of up to 1.0 mM improved the TOCremoval efficiency, perhaps by decreasing the suiphide inhibition of the methanogenicbacteria. Stripping hydrogen sulphide from the reactor liquor at 55 °C helped to promotethe treatment efficiency to a maximum of 39% and was effective in lowering the suiphidelevels. Similar to the 35 °C study, suiphide removal by gas stripping appeared to be themost effective means of sulphur management for the thermophilic experiments.Very high acetate concentrations and minute gas production rates were recordedthroughout the experimental program. Since the experiments resulted in the variouscombinations of high and low levels of both sulphate and suiphide, this studyilldemonstrated that compounds in addition to sulphate and suiphide inhibited the methaneproducing bacteria from using acetate. Various wood extractives and chelating agents aresuspected since they are present in high concentrations in this effluent and they are knownto inhibit the anaerobic wastewater treatment microorganisms.ivTable of ContentsPageAbstract iiTable of Contents vList of Tables viiiList of Figures xiAcknowledgment xiiChapter 1 Introduction 1Objective 3Organization of This Thesis 4Chapter 2 Anaerobic Treatment of Pulp and Paper IndustryWastewaters2.1 Overview 52.2 General Background 52.3 The CTMP Process 72.4 Characteristics of CTMP Effluents 82.5 Process Description of Quesnel River Pulp 13Chapter 3 Anaerobic Wastewater Treatment3.1 Overview 153.2 Introduction 153.3 Intermediate Compounds 203.4 Anaerobic Compared to Aerobic TreatmentProcesses 263.5 Considerations in Anaerobic TreatmentProcess Design 313.6 Two Phase Anaerobic Treatment 433.7 Reactor Temperature 50Chapter 4 The Effects of Sulphur Compounds in AnaerobicWastewater Treatment Systems4.1 Overview 564.2 Introduction 564.3 The Sulphur Reducing Bacteria 57V4.4 Sulphur Forms 604.5 Why is Sulphide a Problem? 664.6 Benefits of Hydrogen Suiphide 724.7 Options for Sulphur Management 74Chapter 5 Experimental Materials and Methods5.1 Overview 765.2 Inocula 765.3 Feedstock 775.4 Apparatus 785.5 Analyses 815.6 Duration and Sampling of Runs 825.7 Analytical Procedures 835.8 Plan of Experiments 88Chapter 6 Characterization of the BCTIvIP/TMP Effluent 91Chapter 7 Sulphur Management Strategy 1:Elevated Reactor pH7.1 Overview 947.2 Background 947.3 Results: Effect of Acid Phase Reactor pHat35°C 977.4 Results: Effect of Acid Phase Reactor at 55 °C 1127.5 Conclusions: Effect of Acid Phase Reactor pH 123Chapter 8 Sulphur Management Strategy 2:Inhibition of the Sulphur Reducing Bacteria8.1 Overview 1248.2 Background 1248.3 Results: Effect of Molybdate Addition to theEffluent at 35 °C 1278.4 Results: Effect of Iron and MolybdateAddition at 35 °C 1398.5 Results: Effect of Molybdate Addition to theEffluent at 55 °C 1488.6 Conclusions: Effect of Molybdate Addition 158Chapter 9 Sulphur Management Strategy 3:Removal of Sulphide From Solution9.1 Overview 1609.2 Background 1609.3 Results: Effect of Hydrogen SulphideStripping at 35 °C 164vi9.4 Results: Effect of Hydrogen SuiphideStripping at 55 °C 1789.5 Conclusions: Effect of Hydrogen SuiphideStripping 187Chapter 10 Effect of Temperature, Time and PhaseSeparation10.1 Overview 18910.2 Effect of Temperature (35 °C versus 55 °C) 18910.3 Start-up After 40 Days 20410.4 Effect of HRT 20610.5 Effect of Phase Separation 20710.6 Reproducibility of Results 210Chapter 11 Conclusions 212Chapter 12 Recommendations 214List of Acronyms 217References 218Appendices:Appendix 1: Carbon Balance 236Appendix 2: Sulphur Balance 244viiList of TablesTable Page1 Sulphur Management Experiments Performed at 35 °C 892 Sulphur Management Experiments Performed at 55 °C 903 Characterization of the BCTMP/TMP Effluent 934 Effect of Acid Phase Reactor pH on Sulphate Reduction at 35 °C 985 Effect of Acid Phase Reactor pH on Total Dissolved SuiphideConcentration at 35 °C 1016 Effect of Acid Phase Reactor pH on TOC Removal at 35 °C 1037 Effect of Acid Phase Reactor pH on Gas Production Rate at 35 °C 1058 Acid Phase Reactor and Methane Phase Reactor pH at 35 °C 1119 Effect of Acid Phase Reactor pH on Carbonate Consumptionat35°C 11210 Effect of Acid Phase Reactor pH on Sulphate Reduction at 55 °C 11311 Effect of Acid Phase Reactor pH on Total Dissolved SuiphideConcentration at 55 °C 11412 Effect of Acid Phase Reactor pH on Sulphite Concentrationat55°C 11613 Effect of Acid Phase Reactor pH on Thiosulphate Concentrationat55°C 11714 Effect of Acid Phase Reactor pH on TOC Removal at 55 °C 11815 Effect of Acid Phase Reactor pH on Gas Production Rate at 55 °C 11816 Acid Phase Reactor and Methane Phase Reactor pH at 55 °C 12217 Effect of Acid Phase Reactor pH on Carbonate Consumptionat 55 °C 12318 Effect of Molybdate Addition on Sulphate Reduction at 35 °C 12819 Effect of Molybdate Addition on Total Dissolved SuiphideConcentration at 35 °C 12920 Effect of Molybdate Addition on TOC Removal at 35 °C 13221 Effect of Molybdate Addition on Total Gas Production Rateat35°C 13422 Effect of Molybdate Addition on Reactor pH at 35 °C 13823 Effect of Molybdate Addition on Carbonate Consumption at 35 °C 13924 Effect of Molybdate and Iron Addition on Sulphate Reductionat35°C 14125 Effect of Molybdate and Iron Addition on Total DissolvedSuiphide Concentration at 35 °C 14226 Effect of Molybdate and Iron Addition on Suiphite Concentrationat35°C 143viii27 Effect of Molybdate and Iron Addition on ThiosuiphateConcentration at 35 °C 14328 Effect of Molybdate and Iron Addition on TOC Removal at 35 °C 14529 Effect of Molybdate and Iron Addition on Gas Production Rateat35°C 14630 Effect of Molybdate and Iron Addition on Reactor pH at 35 °C 14731 Effect of Molybdate and Iron Addition on Carbonate Consumptionat35°C 14732 Effect of Molybdate Addition on Sulphate Reduction at 55 °C 14933 Effect of Molybdate Addition on Total Dissolved SuiphideConcentration at 55 °C 14934 Effect of Molybdate Addition on Suiphite Concentration at 55 °C 15135 Effect of Molybdate Addition on Thiosuiphate Concentrationat55°C 15136 Effect of Molybdate Addition on TOC Removal at 55 °C 15337 Effect of Molybdate Addition on Total Gas Production Rateat55°C 15338 Effect of Molybdate Addition on Reactor pH at 55 °C 15739 Effect of Molybdate Addition on Carbonate Consumption at 55 °C 15840 Effect of Hydrogen Suiphide Stripping on Total DissolvedSulphide Concentration at 35 °C 16641 Effect of Hydrogen Sulphide Stripping on Sulphate Reductionat35°C 16842 Effect of Hydrogen Sulphide Stripping on Sulphite Concentrationat35°C 17043 Effect of Hydrogen Suiphide Stripping on ThiosulphateConcentration at 35 °C 17044 Effect of Hydrogen Sulphide Stripping on TOC Removal at 35 °C 17245 Effect of Hydrogen Suiphide Stripping on Gas Production Rateat35°C 17346 Effect of Hydrogen Sulphide Stripping on Reactor pH at 35 °C 17747 Effect of Hydrogen Suiphide Stripping on Carbonate Consumptionat35°C 17848 Effect of Hydrogen Sulphide Stripping on Total DissolvedSuiphide Concentration at 55 °C 17949 Effect of Hydrogen Sulphide Stripping on Sulphate Reductionat55°C 18050 Effect of Hydrogen Suiphide Stripping on Sulphite Concentrationat55°C 18151 Effect of Hydrogen Sulphide Stripping on ThiosulphateConcentration at 55 °C 18152 Effect of Hydrogen Sulphide Stripping on TOC Removal at 55 °C 18253 Effect of Hydrogen Sulphide Stripping on Gas Production Rateat 55 °C 18354 Effect of Hydrogen Suiphide Stripping on Reactor pH at 55 °C 186ix55 Effect of Hydrogen Suiphide Stripping on Carbonate Consumptionat55°C 18756 Effect of Sulphur Management Strategies and ReactorTemperature on Sulphate Reduction 19057 Effect of Sulphur Management Strategies and ReactorTemperature on Total Dissolved Sulphide Concentration 19158 Effect of Sulphur Management Strategies and ReactorTemperature on Total Organic Carbon Removal 19459 Effect of Sulphur Management Strategies and ReactorTemperature on Mean Total Gas Production Rate 19560 Effect of Sulphur Management Strategies and ReactorTemperature on Acetic Acid Concentration 19961 Effect of Sulphur Management Strategies and ReactorTemperature on Formic Acid Concentration 20062 Effect of Sulphur Management Strategies and ReactorTemperature on Butyric Acid Concentration 20163 Effect of Sulphur Management Strategies and ReactorTemperature on Propionic Acid Concentration 20264 Effect of Sulphur Management Strategies and ReactorTemperature on Lactic Acid Concentration 20365 Effect of Phase Separation on TOC Removal, Sulphate Reductionand Sulphide Concentration for a 0.6 day HRT at 35 °C and 55 °C 20966 Reproducibility of Results: Repeats of APR pH 7.5 ControlsThroughout the Experiments at 35 °C 21167 Carbon Balance for Acid Phase Reactor pH Experiments at 35 °C 23768 Carbon Balance for Acid Phase Reactor pH Experiments at 55 °C 23869 Carbon Balance for Molybdate Addition Experiments at 35 °C 23970 Carbon Balance for Iron and Molybdate Addition Experimentsat35°C 24071 Carbon Balance for Molybdate Addition Experiments at 55 °C 24172 Carbon Balance for Gas Stripping Experiments at 35 °C 24273 Carbon Balance for Gas Stripping Experiments at 55 °C 24374 Sulphur Balance for Acid Phase Reactor pH Experiments at 35 °C 24575 Sulphur Balance for Acid Phase Reactor pH Experiments at 55 °C 24676 Sulphur Balance for Molybdate Addition Experiments at 35 °C 24777 Sulphur Balance for Iron and Molybdate Addition Experimentsat35°C 24878 Sulphur Balance for Molybdate Addition Experiments at 55 °C 24979 Sulphur Balance for Gas Stripping Experiments at 35 °C 25080 Sulphur Balance for Gas Stripping Experiments at 55 °C 251xList of FiguresFigure Page1 Simplified Biochemistry of Anaerobic Digestion 162 Experimental Apparatus 793 Effect of Acid Phase Reactor pH at 35 °C on Volatile Fatty Acids 1074 Effect of Acid Phase Reactor pH at 35 °C on Volatile Fatty Acids 1215 System Recovery Following 1.0 mM Molybdate Addition to theFeedstock 1316 Effect of Molybdate Addition at 35 °C on Volatile Fatty Acids 1357 Effect of Molybdate Addition at 55 °C on Volatile Fatty Acids 1558 Effect of Gas Stripping at 35 °C on Volatile Fatty Acids 1759 Effect of Gas Stripping at 55 °C on Volatile Fatty Acids 18510 Start-up of the Two Phase System at 35 °C:o42,2 and TOC vs Time and HRT 20511 Recommended Treatment System 215xiAcknowledgmentA number of individuals greatly assisted me with this work. I gratefullyacknowledge their generous donations of both time and talent.Thanks for the financial support from the Science Council of British Columbia, thePulp and Paper Research Institute of Canada (PAPRICAN), the Natural Sciences andEngineering Research Council (NSERC) and the UBC Department of ChemicalEngineering. The Quesnel River Pulp Company provided the liquid effluent and thanks aredue to Mr. Roger Puar in particular for his help and his co-operation. Dr. RichardKerekes, the Director of the UBC Pulp and Paper Centre, and Dr. Sheldon Duff of theDepartment of Chemical Engineering, both deserve special mention for their support.Sincere thanks go to Dr. Richard Branion and Dr. Ken Pinder, my thesis cosupervisors, for their support, advice and friendship. I also gratefully acknowledge thehelpful discussions with the rest of my Review Committee: Dr. Jamie Piret, Dr. DennisOuchi and Mr. Jim Wearing. The sizable contributions made to this work by the skilledworkshop and support staffs of the Department of Chemical Engineering, the UBC Pulpand Paper Centre and the UBC Environmental Engineering laboratory are alsoappreciated.Most of all, I’m grateful to Sandra and our children, Jeremie and Nicolas, for theirsupport, for their patience and for their care. It is to Sandra that I dedicate this effort.xli• life is indeed darknesssave when there is urge,And all urge is blindsave when there is knowledge,And all knowledge is vainsave when there is workAnd all work is emptysave when there is love;And when you work with loveyou bindyours4f to yourseand to one another, and to GodKahlil GibranxliiChapter 1IntroductionAnaerobic digestion is a key route for the decomposition of organic matter innature. The resulting methane formation occurs in such diverse microbial habitats asgarden soil, swamps, rice paddies, marine and fresh water sediments, the wet wood ofliving trees, thermal springs, the rumen of cows, the gut of termites, the gastrointestinaltract of humans and of many other animals, and in waste processing digestors (Mah,1983). By controlling the conditions under which this microbial ecosystem operates, theprocess can be accelerated and thus become a useful wastewater treatment technology.Anaerobic digestion can simultaneously reduce the pollution strength of an effluentstream and generate significant amounts of methane gas. For certain industries, the energyvalue of the wastewater constitutes a substantial fraction of their total energy consumptionand efforts to recover the fuel value of the gas are warranted. For other industries, simplydecreasing the aeration costs and the sludge handling costs, which would be incurredduring aerobic treatment, by means of an anaerobic pre-treatment stage would be of neteconomic benefit.Legislation to protect the environment and cost trimming measures have promotedresearch activity into the industrial applications of anaerobic treatment systems. In thepast, energy was cheap, wastes were considered to have an insignificant value, anddisposal options were many. Economics, attitudes and opportunities have changed.Methane fermentation has been used for municipal and industrial sludgestabilization for over a century. McKinney commented in 1962 that “anaerobic digestion isthe uncharted wilderness of sanitary engineering”. Since that time, tremendous researchefforts have yielded significant gains in the knowledge of the microbiology andbiochemistry of anaerobic wastewater treatment processes. In spite of these advances,1anaerobic processes have been largely dismissed for industrial wastewater treatmentapplications. Reported extreme sensitivity and low rates of organics removal have resultedin the specification of huge and costly vessel volumes.Anaerobic treatment applications have broadened due to a number ofdevelopments in the control of biomass and of the environmental factors which enhancethe efficacy of the process. The anaerobic treatment option bears consideration for warm,non-toxic effluent streams which contain high concentrations of organic compounds yetmay be limited in nitrogen and phosphorus. This process is especially attractive where themethane gas can usefully be burnt on site as a fuel. Existing biological aerobic treatmentfacilities which are in need of upgrading due to production increases or tighteningdischarge regulations could profit by integrating anaerobic treatment systems into theiroperation.Process control remains a significant stumbling block in the operation of anaerobictreatment systems. One of the more promising innovations of these systems is theemployment of two phase operation, where the acid producing bacteria are contained in aseparate reaction vessel upstream of the methane producing bacteria. Two phaseanaerobic treatment systems appear to offer important advantages for the treatment of twoextreme sorts of wastewater. For easily degraded and high strength substrates, theseparation of the acid producing phase from the acid consuming phase facilitates processcontrol and consequently enhances the system stability. On the other hand, for wastestreams such as arise from pulp mills, a two phase anaerobic fermentation system providesthe opportunity for upstream removal of constituents which would otherwise inhibit themethane forming bacteria. In both sorts of effluents, the advantages of process controlpromise to outweigh the disadvantage of added process complexity.2ObjectiveThe objective of this work was to enhance the anaerobic treatment of a mechanical(BCTMP/TMP) pulping effluent by manipulating the concentrations of sulphurcompounds in the reactors. This work centered on comparing the most promising meansof sulphur management which were deemed to be reactor pH, addition of molybdenum,and gas stripping of hydrogen suiphide. Experiments were conducted which employed twophase, continuously fed, high rate digesters for a range of hydraulic retention times. Thesystem was operated at both mesophilic and thermophilic temperatures.3Organization of This ThesisChapter 2 reviews the difficulties in treating effluents produced by the CTMP process.Chapter 3 covers the biochemical basics and the practical aspects of anaerobic treatment.Chapter 4 reviews the difficulties of anaerobically treating sulphur rich effluents andoutlines the options for sulphur management.Chapter 5 describes the apparatus, the experimental protocol, the analytical proceduresand the experimental plan.Chapter 6 reviews the characteristics of the BCTMP/TMP effluent which was usedthroughout the experiments.Chapter 7 presents the rationale for pH control of suiphide inhibition in anaerobictreatment systems. The results for manipulating the acid phase reactor pH at 35 °C and55 °C experiments are presented.Chapter 8 presents the rationale for inhibiting the sulphur reducing bacteria in anaerobicdigestion. The results for the experiments investigating molybdate addition to the effluentat 35 °C and 55 °C are summarized and interpreted.Chapter 9 presents the rationale for removing sulphide from anaerobic reactors. Theresults for stripping the methane phase reactors at 35 °C and 55 °C are presented.Chapter 10 compares the effect of temperature on the major treatment parameters for thethree sulphur management strategies. It also presents the start up data and summarizes theeffects of HRT and phase separation for the experiments.Chapter 11 draws conclusions and makes recommendations for further study.4Chapter 2Anaerobic Treatment of Pulp and Paper IndustryWastewaters2.1 OverviewSection 2 presents a general background of the difficulties posed by the effluentswhich are produced by the pulp and paper industry.Section 3 very briefly describes the process of producing CTMP pulp.Section 4 describes the characteristics of CTMP and TMP effluents, in particularthe compounds which are inhibitory to the wastewater treatment microorganisms.Section 5 is a brief process description of Quesnel River Pulp.2.2 General BackgroundThe pulp and paper industry is a major user of resources and also a significantsource of pollutants. Springer (1986) estimated that the Canadian pulp and paper industryis responsible for half of all waste discharged into our waters.The pulping of wood, either by chemical or mechanical means, results in thesolubilization of the wood constituents and this produces an effluent containing celluloses,hemicelluloses, lignins, tannins and wood extractives which are potentially deleterious toaquatic life. Untreated, a direct discharge of such effluents would cause a rapid depletionof dissolved oxygen and impart toxicity that would make the receiving waters intolerableto aquatic life and unfit for consumption.External pollution control measures are employed in order to remove oxygendemand, suspended solids and toxicity from the effluent streams of pulp and paper mills.While water closure is an emerging technology of some new pulp mills, effluent free5operation is a distant goal for most mills and they will continue to require some sort of endof pipe treatment.Biochemical oxygen demand (BOD) and toxicity removal from pulp mill effluentshave historically been accomplished by means of aerobic treatment processes. Suchsystems include aerated stabilization basins and activated sludge plants. These methodsachieve high BOD and total suspended solids (TSS) removal but they are expensive tooperate, principally due to the energy requirements for the extended periods of aeration inorder to achieve detoxification.With the tightening of regulations which govern effluent discharge from pulp andpaper mills, the existing aerobic wastewater treatment systems may be undersized.Anaerobic pre-treatment is an obvious candidate for adding to the treatment capacity inorder to treat these effluents.Mechanical pulping effluents have generally been considered to be too voluminous,too dilute, too recalcitrant and too toxic to be cost effective for treatment by anaerobicmeans. However, the more recent in-plant measures such as stream segregation, decreasedwater consumption and increased process water recycle have resulted in moderate to highstrength waste streams. Although these effluents are too dilute for chemical recovery to beprofitable, they are too concentrated for conventional aerobic biological treatment. Manyof these effluents, especially the mechanical pulping effluents, appear to be well suited toanaerobic treatment according to a number of investigators (Orivuori, 1985, Fromson etat., 1986, Dinsmore, 1987, Rintala and Vuoriranta, 1988, Lee et at., 1989). These wastestreams are warm (45 °C to 80 °C), are low in suspended solids, and contain significantconcentrations of the low molecular weight organic acids, alcohols and sugars which arereadily metabolized by the anaerobic bacteria (Endo and Toya, 1985, Lee et al., 1989).Conversely, they also contain a significant fraction of toxic, inhibitory and recalcitrantcompounds, presenting a challenge for their removal.6Anaerobic treatment systems have operated effectively in the food processingindustry but their performance in the pulp and paper industry has been mixed. Among theCanadian pulp mills there are four anaerobic treatment systems: Quesnel River Pulp,Sturgeon Falls, Lake Utopia and Stone Consolidated. However, in the McCubbinConsultants Inc. 1992 report “Best Available Technology for the Ontario Pulp and PaperIndustry”, the authors omitted any significant discussion of anaerobic treatment for twomain reasons. First, they did not consider the technology to be sufficiently mature topredict costs. Second, the difficulty of growing biomass for start-up or for recoveryfollowing upset posed a risk which was deemed unacceptable. At its present state ofdevelopment, Pichon et a!. (1986), Maat and Habets (1987) and McCubbin ConsultantsInc. (1992) proffered anaerobic treatment not as an alternative but rather as a means ofpotentially reducing the operating costs of conventional aerobic treatment systems.2.3 The CTMP ProcessChemithermomechanical pulping (CTMP) augments the thermomechanical pulping(TMP) thermal and mechanical stages with chemical addition. CTMP is produced byatmospheric steaming and impregnating the wood chips with a 1% to 5% sodium sulphitesolution under alkaline conditions to further soften the wood chips prior to TMPprocessing. The TMP process involves steaming chips at temperatures up to 140 °C underhigh pressure and then shearing the chips to fibres by means of a disc refiner.The CTMP sodium sulphite chip pre-treatment results in a pulp of low extractivescontent, which is stronger and of better quality than TMP. It also lowers the yield and thusincreases the discharge of organic compounds in the effluents. Sodium hydrosuiphite,dithionite and hydrogen peroxide are the brightening agents commonly used formechanically produced pulp.72.4 Characteristics of CTMP EffluentsCTMP effluents are highly concentrated wastes. Thus the power requirements forsupplying oxygen to aerobic treatment are also high. According to Fromson et at. (1986),a major obstacle to increased CTMP implementation is the problem of effluent treatmentto a level suitable for discharge to the- environment. In a comparison of effluent treatmenteconomics, Pearson (1985) and Fromson et a!. (1986), found that CTMP effluenttreatment costs approximate 5% of the capital and operating costs. This is roughly doublethe cost for kraft mill effluent treatment for the same pulp production capacity (Malinen eta!., 1985).Compared with other pulp mill effluents, CTMP effluents are relatively hot (45 to70 °C), toxic to fish and to bio-treatment microorganisms, and concentrated in bothdegradable and more recalcitrant organic compounds. While CTMP pulp yields rangefrom 80 to 90%, the 10 to 20% of the wood charged to the process which is notrecovered as pulp will appear in the wastewater since there are no solids or chemicalrecovery systems in most CTMP mills.Relatively large amounts of organic materials are dissolved in the CTMP processduring the impregnation, refining and brightening stages (Andersson et a!., 1985). Theharsh pre-treatment conditions of CTMP processes such as alkaline processing, hightemperatures and pressures, long retention times, elevated sulphite levels, peroxidebrightening and water reuse all contribute to increasing the concentration of water solublecompounds which are extracted from the pulp fibre, raising the organic loads in theeffluent (Malinen et a!., 1985, Pichon et a!., 1986).Effluent concentrations vary widely due to wood species and storage conditions,process conditions such as refiner pH, temperature and pressure (Urbantas et at., 1985),pulp yield, the dosage and type of bleaching chemicals, pulp washing, product brightness,product freeness and the degree of process water closure (Virkola and Honkanen, 1985,Dinsmore, 1987, Gunnarsson et a!., 1989 and Jackson, 1989).8Under the slightly acidic, high temperature, high pressure conditions of the TMPprocess, simple carbohydrates are the principal components extracted. Lignin is solubilizedto only a very limited extent (Sierra-Alvarez et aL, 1990). Conversely, the alkaline pulpingconditions of CTMP processes are more effective at attacking both the hemicellulose andlignin fractions. While.this improves the quality of the pulp produced, it. also lowers theyield and more than doubles the discharge of organic compounds relative to TMP effluents(Beak Consultants, 1986). This gives rise to the opaque brown coloured effluents whichare characteristic of CTMP processes.Comparing TMP and CTMP effluents, the following characteristics can beascribed directly to the use of sodium suiphite:1) the resin acid concentration of CTMP effluents is double that of TMP effluents(Thakore eta!., 1977, Servizi and Gordon, 1986),2) fatty acids are released into CTMP effluent whereas only very small fatty acidconcentrations are present in TMP effluents (Bennett et a!., 1988),3) the soluble BOD5 is nearly tripled compared to TMP (Bennett eta!., 1988), and4) the effluent is high in unreacted sulphite.The high strength CTMP effluent contains pulp fibres, sand, dirt, pulphemicelluloses, starch, lignin, fatty acids, ethanol, methanol, triglycerides, resin acids andother wood extractives as well as assorted sulphur compounds.The principal contributors to the COD of CTMP effluents are lignins and ligninderivatives (30 to 40% of the total COD), organic acids (35 to 40%) and carbohydrates,present either as pulp hemicelluloses or as starch (10 to 15%) (Pichon et a!., 1987, Hall,1987). Jarvinen et a!. (1980) assessed the composition of TMP effluents. They measured40% each of lignin and carbohydrates with wood extractives being the balance of thedissolved organic content. The elevated lignin and acetate loads in the process water aredue to the alkaline conditions employed by CTMP processing in order to achieve a highfibre recovery (Virkola and Honkanen, 1985, Sierra-Alvarez eta!., 1990).9Lignin is generally considered to be recalcitrant to anaerobic or aerobicbiodegradation under retention times of practical interest. Sierra-Alvarez et at. (1990)reported that the lower molecular weight lignins demonstrated inhibitory effects towardsthe methanogens. Given the 30 to 40% lignin portion of the COD, the treatment efficiencywill be limited to 60 to 70% (Cornacchio and Hall,. 1988). Lower COD removals are.commonly reported. This range is mirrored in the COD:BOD5 ratio reported in theliterature for CTMP effluents, from 2.0:1 to 3.3:1 (Andersson et a!., 1985, BeakConsultants, 1986). Andersson et al. (1985) determined the B0D28 to be approximately70% of the COD, indicating the limits of biodegradability.Inhibitory Compounds in CTMP EffluentsThe nature of the compounds which both impart acute toxicity to fish and inhibitthe anaerobic microorganisms in the waste treatment system in CTMP effluents is welldocumented (Sierra-Alvarez eta!., 1993). The following compounds are generally deemedresponsible:a) wood extractives such as resin and fatty acids, tannins and lignins,b) bleaching chemicals and oxidants such as hydrogen peroxide,c) chelating agents, such as DTPA (diethylene triamine penta acetic acid), ande) inorganic sulphur compounds such as sulphide, sulphite, thiosulphate and sulphate.The effects of the wood extractives, hydrogen peroxide and DTPA on ananaerobic treatment system are discussed below. The effects of the oxidized sulphurspecies are reviewed in Chapter 4.a) Wood ExtractivesMore than 1000 extractable organic substances have been isolated from effluentsproduced by pulp and paper mills (Wilkins and Panadam, 1987). These are naturallyoccurring components of the wood furnish and are released during the pulping andbrightening processes. Classes of low molecular weight extractable substances which are10commonly detected in pulp mill effluents include resin acids, fatty acids, alcohols andaldehydes.According to Leach and Thakore (1977), the major toxic factors in effluents fromCanadian softwood krafi, sulphite and mechanical pulping operations are the family ofapolar compounds which are known collectively as resin acids. These diterpene tricyclicscontribute to effluent toxicity far out of proportion to their minute concentrations. Theyaccount for 60 to 90% of the overall toxicity of mechanical pulping effluents (Leach andThakore, 1976). They adversely affect both fresh water and marine life with median lethalconcentrations (juvenile rainbow trout, 96hLC50) in the 0.2 to 1.1 mg/l range (Leach andThakore, 1977) and have caused chronic effects at concentrations as low as 0.01 mg/I.Besides resin acids being acutely toxic to fish, a number of investigators (Welander, 1985,Field, 1989, McCarthy et at., 1990, Habets and de Vegt, 1990 and Sierra-Alvarez et at.,1990) also have identified resin acids as one of the classes of compounds responsible formarkedly decreasing the activity of the methane producing bacteria.The resin acid levels in pulping effluents tend to vary markedly within and betweenmills. The resin acid concentration in the effluent, and consequently the toxicity, is afunction of: the wood species, the season and the age of the furnish, the concentration offibres in the effluent, and the pH of the effluent (Swan, 1973, Rogers et a!., 1979,Gunnarsson et a!., 1989, Habets and de Vegt, 1990, Richardson et at., 1990 andO’Connor et at., 1992).b) Hydrogen PeroxideHydrogen peroxide is commonly used to brighten CTMP pulp. The Quesnel RiverPulp Co. hydrogen peroxide system leaves a residual of 50 to 100 mg/l peroxide in thewhitewater prior to any wastewater treatment unit (Beak Consultants, 1986). Theconcentration of peroxide in the effluent depends upon the desired brightness of the pulpand on the brightening system itself11Hydrogen peroxide is a highly reactive oxidizing agent. It will readily oxidizeorganic as well as inorganic compounds such as suiphide or suiphite. The hydrogenperoxide degradation products are oxygen and water. Peroxide decomposition is due toboth biocatalytic action and chemical reactions between the peroxide and reducedcompounds present in the process water (Welander and Andersson, 1985).-- As aconsequence, this degradation is accelerated by the concentration of oxidizeabiecompounds and by the level of free metals in the effluent (Beak Consultants, 1986).Hydrogen peroxide is bactericidal, especially to obligate anaerobes (Morris, 1976).The methanogens, unlike the aerobes and facultative anaerobes, lack catalase which isnecessary for biocatalytic peroxide decomposition. Fortunately, some of theheterogeneous group of the acid forming bacteria which are found in anaerobic treatmentsystem are facultative anaerobes. These microbes, as well as activated sludge, containcatalase and they therefore have the capacity to rapidly degrade peroxide. Welander(1989) found that peroxide levels of up to 200 mg/i were tolerated without disrupting theanaerobic treatment process but increasing the peroxide levels above this concentrationcaused reactor failure. Peroxide detoxification tanks upstream of the anaerobic system,configured as either a two phase anaerobic system or as a vessel containing recycledactivated sludge, reduced the peroxide to tolerable levels according to the excellent workof Welander and co-workers since 1984.c) DTPADTPA, diethylene triamine penta acetic acid, is a cationic chelating agentcommonly employed when CTMP is bleached with peroxide. It is added as a pentasodiumsalt to the processing water at concentrations up to 2000 mg/l (Bambrick, 1985) to bindmetal ions which originate in the wood fibre, process water and equipment. DTPA helpsstabilize hydrogen peroxide against free metal catalyzed decomposition and thus ensuresconsistent and effective peroxide brightening at relatively low cost.12The presence of DTPA in effluent streams leads to downstream wastewatertreatment difficulties for two major reasons:1) DTPA binds certain trace metals such as iron, nickel or cobalt which are micro-nutrients essential for microbial growth. Thus the use of DTPA may compromise themetabolism of the downstream biological treatment microorganisms -by- depriving them ofnutrients.2) Strong chelating agents were shown by Mueller et a!. (1976) and Wilkinson (1967) tobe inhibitory to bacteria. While some investigators describe DTPA as having bactericidaleffects (Andersson et a!., 1985, Welander, 1989), Habets and de Vegt (1990) list DTPAas only a minor toxicant. DeVries (1989) concurred with the latter. He demonstrated thatDTPA dosages of 100 mg/i and 400 mg/i resulted in inhibition of unadapted sludge of only2.5% and 12.4% respectively. No mechanism of inhibition was proven.Kennedy et a!. (1991b) also found that acute exposure to DTPA had little impacton acidogenic and methanogenic activity in both batch and continuous tests. However,chronic DTPA exposure at levels up to 180 mg/i led them to observe a decline of themethanogenic activity, perhaps due to the essential cations being complexed and thereforeunavailable for the growth of methanogens. -Both Kennedy et a!. (1991b) and Driessen and Wasenius (1993) found thatprotection against this nutrient deprivation was achieved by the addition of iron as FeC13which apparently functioned as a sacrificial cation. Kennedy et a!. (1991b) proposed thetheory that DTPA preferentially complexes with iron and thus releases other boundcations, making them available to the microbes.2.5 Process Description of Quesnel River PulpThe Quesnel River Pulp Company is a joint venture of West Fraser Timber Co.Ltd. and Daishowa Canada Co. Ltd.. The mill alternates between TMP and CTMP toproduce nominally 950 ADTPD (air dried tons per day) of pulp. Sodium hydrosulphite is13added to the TMP process and sodium silicate helps to stabilize the hydrogen peroxidebrightening stage. Sodium sulphite is used by Quesnel River Pulp in CTMP chip steamingand impregnation. Hydrogen peroxide is used in order to brighten the pulp. A 50 to 100mg/l hydrogen peroxide residual in the mill’s whitewater system of the CTIVIP line istypically found (Beak Consultants, 1986). DTPA is usually added in excess such thatabout a 100 mg/l uncomplexed DTPA residual is present (Kennedy et a!., 1991b). Waterusage at Quesnel River Pulp is typically 16 and 20 m3/ton for TMP and CTMP pulpingrespectively (Habets and de Vegt, 1990).The existing Quesnel River Pulp wastewater treatment system consists of streamsegregation, rotary screens, primary clarification by both gravity sedimentation anddissolved air flotation supplemented by polyethylene oxide and a phenolic resin to assistsolids removal, a 5000 m3 anaerobic pre-treatment tank, two parallel 3500 m3 upflowanaerobic sludge bed (UASB) reactors followed by a 53,000 m3 aerated stabilization basin(Rankin et a!., 1992), another dissolved air flotation unit and finally discharge to theFraser River, The previous configuration of a primary clarifier and 5 day, two cell, aeratedlagoons with submerged aerators resulted in treated effluents which consistently failed tomeet the Canadian federal regulations for BOD5, TSS and fish toxicity (Servizi andGordon, 1986, Beak Consultants, 1986).14Chapter 3Anaerobic Wastewater Treatment3.1 OverviewSection 2 gives a brief background of the biochemical reactions which compriseanaerobic treatment.Section 3 describes the roles of the important intermediate compounds inanaerobic treatment biochemistry.Section 4 compares anaerobic treatment with aerobic treatment.Section 5 reviews the considerations in anaerobic treatment process design.Section 6 considers the merits of the two phase anaerobic reactor configuration.Section 7 summarizes the attributes of thermophilic versus mesophilic operation.3.2 IntroductionAnaerobic digestion is comprised of a cascade of biochemical conversions in amicrobial food chain where, in the absence of oxygen, a product excreted by one bacterialgroup is consumed by another group. The net result of the complex interactingpartnerships between the many types of the metabolically different bacteria is theconversion of water soluble organic compounds to methane, carbon dioxide and smallamounts of hydrogen, nitrogen, suiphide and bacterial cells. The population diversity andinterspecies interactions, which are characteristic of anaerobic digestion, confer bothstability as well as the ability to metabolize a range of energy sources in wastewaters.Anaerobic decomposition of organic material is generally considered to progress infour stages. They are: hydrolysis, acid production, obligate proton reduction, and methaneproduction. For effluents which contain high concentrations of sulphur compounds,15Carbohydrates,Fats and ProteinsHydrolysisSugars,Fatty Acids andAmino AcidsAcidogenesisOxidized InorganicSulphur Compoundsp.Reduction ofOxidized InorganicSulphur CompoundsVolatile Fatty AcidsObligate ProtonReductionCarbon Dioxide4Reduction4 Decarboxylation andMethyl ProtonationFigure 1: Simplified Biochemistry of Anaerobic DigestionAcetogenesisHydrogen andCarbon DioxideAcetate,Formate and MethanolMethane andCarbon DioxideSulphide andCarbon Dioxide16sulphur reduction also plays a significant role. The sequence of these steps is illustrated inFigure 1. They are briefly reviewed below.HydrolysisIn the first “hydrolytic” stage, complex organic molecules such as polysaccharides,proteins and lipids are broken down by extra cellular enzymes which are produced by theacidogenic bacteria. The products include simple compounds such as sugar monomers,amino acids and alcohols, and long and short chain fatty acids respectively.AcidogenesisThe second stage is known as “aciclogenesis”. The acid forming orchemoheterotrophic, fermentative bacteria which catabolize the products of hydrolysis arecomprised of a mixed microbial population of both obligate and facultative anaerobes.Thus small quantities of dissolved oxygen can be tolerated by the acid forming stage.Gram-negative, nonsporing, obligately anaerobic rods which produce hydrogen andvolatile fatty acids (VFAs) predominate in sewage sludge (Zeikus, 1977). Gram-positive,facultative rods, gram-positive cocci, endospore-forming rods and gram-positiveasporogenous rods are also routinely found. Genera such as Bacteroides, Clostridium,Butyrivibrio, Eubacterium, ByIdobacterium and Lactobacillus are among thepredominant organisms found in anaerobic treatment systems (Toerien and Hattingh,1969).In acidogenesis, the degradation products from the hydrolytic stage are transportedinto the bacterial cells and are fermented to form the C1 to C4 monocarboxylic or shortchain volatile fatty acids, lactic acid, ethanol, methanol, ammonia, hydrogen and carbondioxide. Some of these intermediates such as hydrogen, methanol, formate, acetate andcarbon dioxide are metabolized directly by the methanogenic bacteria but the fatty acidsother than acetate and formate must first be catabolized by the obligate proton reducingacetogens.17The volatile acid metabolites produced in the acid phase reactor can be producedfrom carbohydrates, from some amino acids and from lipids. For practical retention timeshowever, the protein and lipid contribution to the total substrate pool is not significant(Eastman and Ferguson, 1981). Almost all of the products formed in the acid phasereactor can therefore be expected to be derived from the carbohydrate fractionThe idealized reactions of glucose by the fermentative bacteria are as follows withGibb’s free energy in U/mole of reactant (Mclnemey and Bryant, 1981):C61120+ 2CH3C00 -,2CH3COO + 2HC03 + 2HAG° = -302.9C6H120+1120 -*CH32OO- + CH3OO + 11C03 +112+ 3HAG° = -282.4• C6H120+21120 -* 2CH3CO + 211C03 + 2HAG° = -225.9C6H120-,2CH3CHOHCOO + 2HAG° = -198.3Obligate Proton ReductionThe obligate proton reducing acetogenic bacteria dehydrogenate the fermentationproducts other than acetate and hydrogen (three or more carbon fatty acid chains or twoor more carbon alcohols) by a biochemical process known as ‘beta oxidation” (Jeris andMcCarty, 1965). This results in the cleavage of two carbon acetic acid fragments from thefatty acid chain and also produces hydrogen and carbon dioxide. Odd carbon chained acidsare beta oxidized to form acetate, propionate and ethanol (Boone and Mah, 1987).It is only at extremely low hydrogen partial pressures (1 .6 x 16 to 5.8 X 1 Oatm) that these bacteria grow and produce H2 and acetate (Boone and Mah, 1987).Therefore, the H2 producing bacteria require the presence of hydrogen utilizing bacteriasuch as the methanogens or the sulphur reducing bacteria. If hydrogen removal is lessefficient, hydrogen blocks the proton reduction route and the fermentative bacteria18produce reduced fermentation products such as propionate, butyrate, lactate or ethanol(Boone and Mah, 1987). During the hydrolysis and fermentation sequences, little COD isremoved. These processes together consist of catabolizing large molecules into one or twocarbon fragments and producing hydrogen and carbon dioxide as by products. Theproducts which are formed during the hydrolytic, aceto-genic and proton reductive stagesare well suited to the last step of methane formation.MethanogenesisThe final stage of anaerobic digestion produces methane by means of theMethanobacteriaceae, constituted by the following genera: Methanobacterium,Methanosarcina, Methanococcus, Methanospirilium and Methanothrix (Novaes, 1986).Few groupings of microorganisms are as morphologically diverse as the methaneproducing bacteria (MPB). The bacteria may be in the form of sarcina (packets of irregularsized cells), rods, cocci (spheres) and spirals and they often appear as pairs, chains or longfilaments. These cells demonstrate a tendency to adhere onto solid surfaces and toaggregate into clumps, floes or granules.The terminal organisms in the microbial food chain, the MPB appear to be unifiedonly in their common capacity to form methane. All of the MPB are strict anaerobes andare able to utilize only a few substrates. The methanogens are limited to the catabolism ofone or two carbon compounds. All are able to form methane from hydrogen and carbondioxide. Several of these species can use formic acid. Although acetate is the majormethanogenic precursor, only the Methanosarcina species, Methanothrix soehngenii andMethanococcus mazei can form methane from acetate (Mah, 1983). Methanol andmethylamines are also utilized by a few IvIPB and are of limited importance in mostanaerobic treatment systems (Zeikus, 1977, Balch eta!., 1979 and Koch, 1981).Methane is produced by two distinct pathways. In one route, acetate isdecarboxylated to form methane and carbon dioxide. In the other methanogenic pathway,hydrogen is oxidized and carbon dioxide is reduced to form methane and water. It is19widely quoted that acetate accounts for approximately 70% and carbon dioxide accountsfor approximately 30% of the methane carbon formed in anaerobic digesters or in naturalenvironments. Of course, this figure varies with the substrate and the microbial consortiapresent.Methane formation is generally the rate limiting step in anaerobic wastewatertreatment. After system upset or following periods of overloading, it is the methanogenicstage which requires the longest period for reacclimation.The range of organic substrates utilized by the sulphur reducing bacteria (SRB) isfar wider than earlier work suggested and more diverse than for the MPB (Imhoff andPfenning, 1983, Laanbroek and Pfenning, 1981, Postgate, 1979). Lactate, propionate,butyrate, acetate and amino acids are potential substrates to support sulphate reduction(Smith and KIug, 1981).3.3 Intermediate CompoundsThe intermediate compounds of primary importance and their roles in anaerobictreatment biochemistry are listed below.Hydrogen and Carbon DioxideThe reduction of carbon dioxide to methane by molecular hydrogen is the energyyielding reaction common to all of the MPB. Methane is formed by the reduction ofcarbon dioxide according to the following reaction with AG° in KJ/mole of reactant (Jechand Brautigam, 1983):4112+ HC03 + H -* CR4 +31120AG° = -135.6In addition to being an important substrate for the MPB, hydrogen is the electron donorfor sulphate reduction by the SRB. The SRB reduce sulphate and suiphite to suiphide,using the oxidized sulphur compound as the terminal electron acceptor for respiration. TheSRB’s carbon metabolism is generally incomplete, yielding hydrogen (Battersby, 1988),20acetate and carbon dioxide from diverse substrates (Cappenberg, 1972, Ueki et at., 1986,Battersby, 1988 and Hilton and Oleszkiewicz, 1988). The low hydrogen concentrationswhich are detected in digesters likely stems from the rapid hydrogen uptake by both theMPB and the SRB. Sulphate is reduced according to (Mclnerney and Bryant, 1981):4112+ so42-+11 —* HS- +41120AG° =-151.9The large negative free energy associated with both of the hydrogen utilizationreactions indicates that the reaction favours hydrogen use. Together, the MPB and theSRB maintain a low partial pressure of hydrogen in the reactor liquor. The coupling ofhydrogen consumption with VFA catabolism makes these fermentative reactionsthermodynamically favourable. Hydrogen is involved in all of the principle reactions. Itexerts control on the acid formation stages and is a key substrate for the methanogens andthe sulphur reducing bacteria. The VFA catabolizing, hydrogen producing reactionsbecome more favourable at lower hydrogen concentrations.Acetic AcidAcetic acid is a product of the degradation of carbohydrates as well as anintermediate product of the beta oxidation of longer chained fatty acids. Even though onlya few of the methanogenic species have been confirmed to degrade acetate, depending onthe wastewater and environmental conditions, acetate accounts for 60 to 90% of themethane formed in most anaerobic treatment systems (Smith and Mah, 1966, Mountfortand Asher, 1978).The decarboxylation of acetate results in the methyl group of acetate beingreduced to methane and the carboxyl group being oxidized to carbon dioxide as follows(Smith et at., 1980 and Jech and Brautigam, 1983):CH3OO +1120 -* CR4 + HC03AG° = -31.021Acetate is also oxidized to carbon dioxide by the sulphur reducing bacteriaDesufotomaculum acetoxidans (Widdel and Pfenning, 1977) and Desu’fobacteriumpostgatei (Widdel and Pfenning, 1981). Complete oxidizing SRB such asDesu’fotomaculum species could be termed “acetoelastic”. Acetate does not support thegrowth of all SRB however. Despite the - significance of- acetate as an intermediatefermentation product, acetate-utilizing SRB have only recently been isolated (Widdel andPfenning, 1981, Ingvorsen et at., 1984), and their population levels and activities inanaerobic treatment systems have not been established (Phelps et a!., 1985). Isa et at.(1986), using well digested anaerobic sludge, Laanbroek and Pfenning (1981),investigating Desz4fobulbus, and Nanninga and Gottschal (1987), experimenting with fourDesulfovibrio species isolated from an anaerobic treatment plant: D. desu’furicans, D.sapovorans, D. carbinolicu and Desusfobulbus propionicus (proposed), found SRBwhich did not metabolize acetate. Other investigators (Phelps et a!., 1985, for example)found acetate as well as formate and methanol to not be used by SRB (Desufovibriovulgaris) alone but was used in co-culture with Methanosarcina barkeri.Acetate appears to be a readily utilizable substrate by SRB isolated from brackishwater and marine mud (Laanbroek and Pfenning, 1981, Widdel and Pfenning, 1981,Battersby, 1988, Heinrichs et a!., 1990). Even in freshwater sediments, some investigators(Winfrey and Ward, 1983, Ingvorsen et a!., 1984) have identified acetate to be a majorsubstrate consumed in the course of sulphate reduction. Laanbroek and Pfenning (1981)could only isolate these acetate-oxidizing SRB strains from marine or brackishenviromnents. Such bacteria oxidize two moles of acetate to two moles of carbon dioxidefor each mole of S042 reduced to 2 according to Hilton and Oleszkiewicz, (1987a):2CH3COOH + S042 > S2 + 2C0 + 4H20Acetate does not usually accumulate in anaerobic environments because it servesas a major energy source for some species of both the SRB and the MPB. At the bottomof the food chain, the acetoclastic SRB are in direct competition with the MPB for acetate22and hydrogen. Potentially, the SRB are capable of oxidizing acetate directly to carbondioxide (Hilton and Oleszkiewicz, 1987b), thus diverting a considerable fraction of themethane potential of the wastewater from the methanogens.Formic AcidFormate is a product of many fermentations. Since it is readily cleaved intohydrogen and carbon dioxide (Zeikus, 1977), formate and hydrogen are nearly equivalentenergy sources for methanogens. Formic acid is one of the few metabolites which isdirectly metabolized by many of the methanogenic bacteria. Hence, its formation in theacid phase reactor is an asset from the standpoint of potential methane production. Thereaction is written as (Pipyn and Verstraete, 1981):4HC02 + H20 + H -* CR4 + 3HC0AG° = -134.3Formic acid uptake in the methane phase reactor is reportedly very rapid (de Haastet a!., 1986). This appears to account for the very small concentrations reported in theliterature (Hall and Adams, 1986).Propionic AcidPropionic acid is formed from the breakdown of odd numbered long carbon chainvolatile fatty acids. Methane formation from propionic acid is considerably slower thanfrom acetate or n-butyrate (Hall and Adams, 1986). Consequently, the reactor conditionssuch as a pH of 6.0 to 7.0 and high hydraulic loading rates which favour propionic acidformation should be avoided (Eastman and Ferguson, 1981).Pipyn and Verstraete (1981) gave the anaerobic catabolic oxidation of propionicacid to form acetate, hydrogen and carbon dioxide as:CH32OO +21120 -C11300 + + 3H2 + CO2AG° = -76.1From elementary chemistry principles, it is easy to see that this delicate reactionwill be directly inhibited by high concentrations of either acetate (Eastman and Ferguson,231981) or hydrogen (Bull et at., 1984) and indirectly inhibited by other long chained fattyacids since their breakdown products are also acetate and hydrogen. Increased propionateis the consequence of overloading or artificially raised hydrogen concentration (Asinari diSan Marzano, 1981, Mackie and Bryant, 1981, Kaspar and Wuhrmann, 1978).In anaerobic reactors which are fed propionate, the conversion of sulphate tosuiphide is efficient, leading to large concentrations of sulphide. Propionate-oxidizing SRBstrains, the Desuifobulbus, were found in both freshwater and marine sediments (Pfenninget at., 1981). Propionate is produced from long chain VFAs which have odd numbers ofcarbon atoms. Propionate breakdown is mediated by bacteria which fulfill twocomplementary roles: a hydrogen producing acetogenic bacteria and a hydrogen utilizingbacteria.n-Butyric AcidWhile acetic acid is the intermediate of major importance in methane formation, nbutyric acid is a dominant VFA metabolite in the acidogenesis of carbohydrates underacidic conditions (Andrews and Pearson, 1965). n-butyric acid is converted to acetateaccording to the reaction (Jech and Brautigam, 1983):CH32OO +21120 -* 2CH3C00 + 2112+ HAG° = -48.1This butyric acid degradation reaction is accelerated by the presence of hydrogenutilizing methanogens (Hall and Adams, 1986) or by SRB. Butyric acid is also used by theSRB according to (Mclnerney and Bryant, 1981):2CH3COO + S042 —> 4CH3C00 + 11S +AG° = -55.7The products of the reaction are suiphide and the primary substrates of themethane forming bacteria: acetate and hydrogen. Under sulphate limiting conditions, SRBplay an acetogenic role and the MPB act as the terminal electron acceptor (Phelps et a!.,1985, Battersby, 1988).24MethanolMethanol can be converted to methane through at least two routes: directly(Zeikus, 1977) and via intermediate formation of VFAs (Lettinga et at., 1979).4CH3011 -* 3CH4 + CO2 + 2H0AG° = -314.8In anaerobic environments, methanol is produced in only very small quantities.Unless methanol is already present in the wastewater (as it is in some pulp mill effluentsfor example), the formation of methane from methanol is of little consequence. Methanolis also utilized by the SRB.EthanolElevated ethanol concentrations in the acid phase reactor have been linked to bothshock loading and to high organic loads (Li et at., 1984, Wildenauer and Winter, 1985).Ethanol is formed from glucose as follows (Pipyn and Verstraete, 1981):C61120+21120-> 2C113C10H+ 2HC03 + 2HAG° = -225.9Ethanol is in turn degraded to acetate and hydrogen according to the followingreaction (Pipyn and Verstraete, 1981):2CH3CO + H20 -> 2C113C00 + H20AG° = +9.6Lactic AcidLactic acid, like ethanol, has been implicated as an indicator of excessive organicloading rates, but it disappears once the balance between the hydrogen producingacidogens and the hydrogen utilizing MPB or SRB is restored (Andrews and Pearson,1965). In the acidogenic stage of anaerobic treatment, two moles of lactic acid are formedper mole of glucose as follows (Jech and Brautigam, 1983).C61120+21120 -*2CU3CHOHCOO + 2HAG° =-198.325The conversion of lactate to acetic acid is favoured under conditions of lowhydrogen concentrations. The reaction is very fast and is written as (Pipyn and Verstraete,1981):CU3HOHCOO +21120 -, C11300 + HC03 + H +2112- AG°=-4.2Under elevated hydrogen concentrations, propionate formation from lactate can beexpected to be favoured (Hall and Adams, 1986).In the presence of sulphate, lactate is also used by the SRB to produce acetate. Anexample of such “acetogenic” SRB are the Desu41fovibrio species which incompletelyoxidize two moles of lactate to two moles of acetate and two moles of bicarbonate foreach mole of sulphate reduced to sulphide. Mclnerney and Bryant (1981) list thebiochemical reaction as follows:2CH3CHOHCOO + S042 -* 2CU3C00 + HS- + H + 211C03AG° = -160.33.4 Anaerobic Compared to Aerobic Treatment ProcessesBiological wastewater treatment can proceed by means of aerobic or anaerobicmixed populations of microbial cells. Aerobic microbial degradation of the organicconstituents of the wastewaters results in the growth of more bacterial cells and carbondioxide. This is quite different from anaerobic treatment where the mixed microbialpopulation produces methane as the predominant metabolite.A number of factors affect the technical and economic viability of the anaerobicprocess:1) the wastewater composition and concentration which influences the treatment rate andtreatment efficiency, and biomass production,262) the availability of inocula and the ability to grow and retain an active microbiologicaltreatment population,3) the requirements for, and the ability to deliver, nutrients and micro-nutrients such asnitrogen, phosphorus and trace metals to the bacteria,4) the buffering capacity of the effluent and the requirement for pH control solutions, -5) the requirements for upstream processing in order to add or remove heat, or to removesuspended solids or compounds which are inhibitory or toxic to the anaerobic bacteria,6) the stability of the system to variations in wastewater composition and volume,7) the formation of metabolites which impart end-product inhibition,8) the value of the biogas produced, its composition and scrubbing costs prior to usecompared to its energy equivalent,9) the treatment efficiency needed to satisfy downstream unit operations or to meetdischarge requirements.10) the energy cost for anaerobic treatment versus the energy cost for removing an equaloxygen demand by aerobic means, and11) the cost of nutrients and of excess sludge disposal.Advantages of Anaerobic Treatment SystemsAnaerobic digestion has a number of significant advantages over the morecommonly encountered aerobic treatment alternatives:1) No power intensive aeration or other significant power input is needed for anaerobictreatment. This is an important attribute for concentrated and toxic effluents which requireextended treatment times prior to discharge.2) The high biomass densities and the absence of a significant mass transfer limitation suchas occurs with oxygen transfer in aerobic processes allow anaerobic systems to toleratehigh organic loading rates (10 to 30 kg COD/rn3 day) (Maat and Habets, 1987) and shockorganic loads (Gunnarsson and Rosen, 1985). This results in very small spacerequirements for anaerobic treatment systems.273) The low rate of cell synthesis, which is characteristic of anaerobic bacteria(approximately 5% of removed COD versus 50% CODr for aerobic systems), results inlower nutrient requirements as well as reduced sludge disposal problems.4) With acclimated anaerobic bacteria, long dormant periods of up to a year are possiblewith a rapid start-up after shutdown (Kennedy et al, 1991a, Lettinga eta!., 1990)..5) Hot (up to 65 °C) effluents may be tolerated without a cooling stage, since acclimatedanaerobes can function in the thermophilic range (50 to 65 °C) as well as in the mesophilicrange (20 to 37 °C).6) 0.35 m3 of methane rich fuel gas (less the cell yield) is produced per kg of CODremoved by the methane forming bacteria. This presents the possibility for displacingpurchased fuel or for earning operating revenues to reduce the overall expense of treatinghigh strength wastewaters rather than bearing the on-going costs of aerobic treatment.The reduced costs for nutrients, sludge disposal and power for aeration are theprincipal factors leading to the inclusion of anaerobic operations into the treatment train(Speece, 1983, Habets and Kneilissen, 1985 and Wilson et a!., 1987). Fromson et a!.,(1986) noted that at almost one kWh/kg BOD removed, the operating cost advantage ofanaerobic over aerobic treatment systems is around $5/ADT of pulp. Rekunen (1985) andMaat and Habets (1987) quoted UASB operating costs as 10% of the activated sludgecosts for the same capacity, while the capital costs for anaerobic systems are of the sameorder of magnitude as for aerobic units (Turk, 1988). These high aerobic operating costscome without any benefit to the industry. Andersson et a!. (1985), Hall (1987) andDinsmore (1987) concluded that high rate anaerobic treatment/aerobic polishing systemsare the most cost effective alternative for high strength, large volume effluents. Thisconfiguration can aid pulp mills in upgrading their treatment systems to satisfy dischargeregulations or to accommodate increased production.28Disadvantages of Anaerobic Treatment SystemsAnaerobic systems have an obstinate reputation of being unreliable and difficult tocontrol and maintain. According to Dinsmore (1987), the major disadvantage of theanaerobic process is its need for stringent process control, approximately equal to thatrequired for activated sludge treatment. However, there are few parameters that needcontrol in anaerobic treatment. These include: the density and contact with substrate ofactive bacterial cells, the organic and hydraulic loading rates, the reactor pH and thereactor temperature. These requirements do not present insurmountable technicaldifficulties and technically simple and relatively inexpensive reactors can be used.The process limitations of anaerobic treatment systems are as follows:1) The substrate removal per unit anaerobic biomass is approximately 70% lower than foraerobic treatment (Lee et a!., 1989). In order to compensate for this inherent shortcomingand to avoid low volumetric conversion rates, the biomass activity per unit reactor volumeneeds to be promoted by maintaining the biomass densities at least three to four times thatfound in aerobic treatment systems and by operating near the pH and temperature optima.2) The low cell synthesis which is an advantage during steady state operation becomes asignificant disadvantage during start-up or recovery from a system upset. Consequently,biomass retention is critical in order to avoid long start-up and recovery times aftertoxification or overloading. However, the control of active biomass is difficult since:i) for suspended growth systems, the active methane producing bacteria are buoyed by thegas to the liquor surface and thus biomass retention necessitates special provisions forgas/liquid/solid separation, andii) for attached growth systems, biomass attachment is not optimally distributedthroughout the reactor volume nor is the bioflim thickness adequately controlled byperiodic sloughing.293) The pH of the reactors needs to be near neutral for the sensitive methanogens, althoughthe more robust acid forming bacteria function effectively from a pH of 8.5 down to pH4.5 (Price, 1985). The sulphur reducing bacteria function from pH 6.5 to greater than 8.5.4) The chemically reduced conditions which are necessary for anaerobic processesproduce explosive, toxic, malodorous and- corrosive compounds such as methane,hydrogen sulphide, mercaptans, ammonia and organic acids.5) The methane forming bacteria are sensitive to oxidants such as oxygen, peroxide andchlorine, to sulphide and sulphite, to wood extractives and to metals (Lee eta!., 1989).6) With only a few exceptions of recalcitrant compounds (lignin or high molecular weighttannins for example) and those which are toxic to fish (various wood extractives such asresin acids), compounds which are susceptible to aerobic decomposition can also betreated anaerobically (Speece, 1983) but at much slower rates. Since lignins and tanninsare major contributors to the colour of pulp mill effluent streams, anaerobic treatment ofsuch effluents will not result in significant colour removal.7) Only a minor reduction in pulp mill effluent toxicity is commonly measured duringanaerobic treatment (Wilson et at., 1987). An additional 3 to 5 days aerobic detoxificationstage is typically required in order to consistently meet the requirements for a non-lethalmill effluent. Anaerobic pre-treatment however reduces the subsequent aerobic HRT andenergy requirements in order to render the discharge non-toxic to fish (Turk, 1988,MacLean et at., 1990).8) Even greater than 90% treatment efficiencies of concentrated wastewaters will result ineffluents which remain unsuitable for discharge to the receiving waters. Aerobic cleanup ofthe anaerobically treated wastewater is generally prescribed. Rather than entirely replacingaerobic wastewater treatment, anaerobic systems can remove substantial amounts ofoxygen demand and thus lower both the capital and the operating costs of an aerobictreatment plant.309) Compared to aerobic processes, relatively few industries have installed anaerobictreatment systems. Consequently, the available pool of full scale experience is very limited.3.5 Considerations in Anaerobic Treatment Process DesignThe rational design and effective process control of anaerobic treatment systemsdepends upon an applied understanding of the factors which promote the activity and thestability of the biotreatment microorganisms. The feeding frequency, means of biomassretention and control, mixing, pH, nutrients, temperature and one versus two phases arethe key process design considerations. Each is briefly discussed below.Batch or Continuous OperationUnfortunately, most of the observations in the literature which describe the effectsof sulphur concentrations on bio-treatment microorganisms are based on batch studies.While these experiments are relatively simple to perform and yield extensive andreproducible data, a direct extrapolation to real continuous wastewater treatment systemsuffers from considerable drawbacks.1) The concentrations of substrates are functions of time, unlike continuous reactorsystems.ii) The HRT and SRT are confounded and therefore the effects of these two key variablescannot be independently ascertained.iii) End products accumulate. This may be of special significance when feedback inhibitionis known to occur. There are two such cases of particular relevance to anaerobic treatmentsystems. In the series of reactions collectively known as acetogenesis, hydrogen is formedand needs to be removed to very low levels in order that the acetogenic reactions remainthermodynamically favourable. The other case involves sulphate reduction to hydrogensulphide, where elevated dissolved sulphide levels have been observed to detract from theoverall sulphate reduction efficiency and methanogenic activity. A result of this end31product accumulation is that generally the reactor pH is not constant over the time of thebatch.Batch culture studies exert primary microbial selection pressure based on thephysiological properties of the microorganisms. In a wastewater treatment plant however,the physical properties of the microorganisms such as mass, size and the electrical chargeof the cells exert secondary selective pressures on the microbial population. It is onlyunder continuous operating conditions, especially at short HRTs or SRTs, that thesesecondary factors, magnified by different growth kinetics of the microorganisms, can playan important role in the selection mechanism (Pretorius, 1987).Retained Biomass ReactorsAnaerobic biomass COD removal activity ranges between 0.8 and 2.0 kgCOD/kgVSSday but it is more commonly found to be between 1.0 and 1.5 kg COD/kgVSS/day (van den Berg et al., 1985). Of course, this activity depends upon the substrateand on the operating conditions of the reactor. This figure is small compared to that ofaerobic treatment microorganisms.A key to efficient anaerobic system performance is to maximize both the activityand the density of the bacteria. Retention of the methanogens in particular is essentialsince they grow approximately an order of magnitude more slowly than their SRB andnon-SRB acidogenic partners. Retaining the microbial populations within the reactorindependent of the flow rate enables the HRT to be minimized. A low HRT or a highvolumetric throughput allows the size and the cost of the digesters to be kept low. Theretained biomass reactors of note, the fixed film reactor and the upflow anaerobic sludgebed reactor, have widely been proposed as the most promising anaerobic treatmentdesigns (Li et al., 1984).Microbial activity is linked to the form of biomass retention, either as suspendedflocs or as immobilized films. The form of biomass retention is a selective pressure on theconstituents of the microbial population and it exerts an influence on the mass transfer32characteristics and on acclimation of the microbes to the reactor environment. Butlin et at.(1956), Isa et at. (1986), Winfrey and Zeikus (1977) and Jensen et at. (1988) notedconsiderable adaptation of MPB to free hydrogen sulphide concentrations with fixedbiomass reactors compared to suspended growth systems,Upflow Anaerobic Sludge Bed ReactorThe upflow anaerobic sludge bed (UASB) reactor retains the microbes within thereactor without an internal support structure. Instead, this reactor design relies on thesuperior settling characteristics of flocs and granules. Consequently, the systemperformance and stability is highly dependent upon the settling properties of themicroorganisms. Unfortunately, the granulation process which is necessary for highlysettleable granules has been demonstrated to be an unreliable phenomenon in pulp andpaper mill effluents (Maat and Habets, 1985).The amount of biomass retained within a reactor varies with the organic loadingrate and with the form of the biomass. In order to maintain high biomass densities inUASB systems, granular biomass must be used. Very high organic loading rates can resultin excessive sludge expansion which can lead to wash out of the most active gas-producing biomass. On the other hand, very low loading rates lead to a small amount ofbed expansion and this presents problems of insufficient contact between the influentstream and the bacteria. Loading rates of UASB systems are typically 3.0 up to 30 kgBOD5/m3/day at 35 °C (Lee et at., 1989).Typically, no external agitation is employed for UASB reactors. The upwardflowing wastewater and the gas which is evolved is generally adequate to provide forsufficient contact of the microbes with the effluent. Internally generated mixing also avoidsmixing hardware which may damage the fragile granules.Biomass occupies about 30% of the reactor volume and the overlying 70%functions as a gas/liquid/solid separator (Henze and Harremoes, 1983). The overlyingquiescent liquid layer which is required for separation of the biogas from the biomass flocs33means that reactor height is poorly utilized. Unfortunately, with this design, thegas/liquid/biomass separation occurs in the upper level of the reactor where the turbulencecaused by gas bubbling is greatest. However, this lack of mixing limits the ability tocontrol the reactor (pH for example) and consequently, a number of UASB systems areoperated with a liquid recycle.Bioflim ReactorAnaerobic bioflims are slimy layers which are attached to inert support surfaces.Almost any inert surface inmiersed in a substrate under conditions which are favourable tobiological growth will be colonized by microorganisms. Anaerobic layers may accumulateup to 4 mm in thickness (van den Berg et a!., 1985) and are white, grey, brown or black incolour. The rate of bioflim formation varies widely. The film thickness varies with anumber of factors:1) the nature of the support surface such as surface roughness, specific surface area,porosity, pore size, hydrophobicity (Rosen and Gunnarsson, 1986),2) the localized substrate concentration,3) the environmental parameters such as temperature, pH and turbulence, and4) the reactor configuration (van den Berg eta!., 1985).Film media provide a surface for bioflim development. It increases the bacterialdensity and more evenly distributes the biomass throughout the reactor without energycosts for mixing. The bioflim is important to stabilize the treatment capacity of a process(Rosen and Gunnarsson, 1986). In continuous flow reactors, the MPB adhere moreeffectively than the SRB to inert support surfaces (Isa et a!., 1986). Thus the lower MPBgrowth rates compared to the SRB are at least partly compensated for by superior biomassretention (Isa et a!., 1986, McCartney and Oleszkiewicz, 1990). Bioflims also retard thegas velocity and facilitate gas/solid separation as gas rises up through the reactor liquor,thus allowing unattached biomass to be retained within the reactor and to further improvebiomass density.34The choice of a suitable fixed film support must be made by balancing therequirement for a high specific surface area in order to provide for substantial film growthwith the need for a high void fraction so that biomass accumulation does not lead toplugging over the long term, leading to excessive pressure drops or by-passing.Random versus regular packing is another design choice to be made. Randompacking would maximize the contact opportunities to enable the rising gas and sludge toseparate. Regular rigid vertical channels would seem to be less adept as a solids/gasseparator but would be considerably less likely to clog (de Haast et at., 1983, van denBerg et a!., 1985, Beak Consultants, 1986). Narrowly spaced vertical packing might offerthe best compromise for media configuration by minimizing the risk of clogging whileefficiently using all of the reactor volume without intense and costly recycling ofwastewater.A bioflim reactor or “anaerobic filter” combines good treatment efficiency at a highor variable loading capacity without the need for specialized sludge separation (Rosen andGunnarsson, 1986). However, high suspended solids levels in the feed or excessivebiomass accumulations pose severe operating difficulties, leading to plugging andchanneling and a loss of active reactor volume. These shortcomings, as well as theextended start up times, loss of active reactor volume and the capital costs posed by thefixed media, have discouraged the installation of such units. However, these problems canbe minimized by: restricting suspended solids entry into the reactor system by means ofefficient solid liquid separation, utilizing two phase operation with control of the fastgrowing acidogenic biomass, and designing for down flow as opposed to upflow operationso that suspended solids are flushed out of the reactor.Rosen and Gunnarsson (1986) advocated the employment of a fixed bed anaerobicreactor for most pulp and paper effluents which are variable and which may containinhibiting compounds. Maillacheruvu et al. (1993) found the upflow anaerobic filter ableto tolerate higher sulphide loads than a suspended growth chemostat. According to Parkin35and Speece (1984), immobilized biomass systems allow for rapid quasi-plug flow elutionof toxic slugs while retaining the biomass, thereby allowing recovery and acclimation tooccur since immobilized biomass substantially reduces the risk of sludge wash out. In spiteof these attributes, the experience with full scale installations in the pulp and paperindustry are limited.Hybrid ReactorThe methane forming bacteria grow both as films on inert surfaces as with thefixed film reactor design, and as suspended fiocs or granules in the upfiow anaerobicsludge bed reactor. Rather than choosing between a single mechanism of bacterialretention, utilizing both means in a hybrid of the UASB and fixed film reactors can betteraccommodate elevated loading rates with a high removal efficiencies and can help tominimize the limitations of each individual design.A hybrid or upfiow anaerobic filter reactor is designed as a sludge bed in the lowersection with its dense and highly active biomass. High bacterial densities and activities areachieved with the suspended flocs and granules of the UASB reactor. The greatestfraction of COD reduction takes place in the lowest portions of the reactor where thegranules accumulate (Rosen and Gunnarsson, 1986). However, suspended biomasssystems require long acclimation periods and exhibit poor tolerance to shock loading(Geller and Gottsching, 1985). Installing an inert film support structure to occupy theupper portion of the reactor can help. Raising the media above the very dense sludge bedor increasing the media voidage throughout the reactor volume reduces some of thehydraulic inefficiencies which are observed with a conventional packed bed (Hall, 1987).The packed zone aids in retaining biomass both by slowing the velocity of sludge buoyedup by the biogas and by developing a fixed and active microbial film, thereby contributingto the total treatment capacity.Some investigators have rejected the use of film media in anaerobic reactors due tothe capital cost (Cocci et a!., 1982). However, bioflims can add significantly to treatment36system performance. They can increase the active biomass density, thus adding to overalltreatment efficiency. They tolerate high variations in organic loading as well as thepresence of toxic substances compared to suspended biomass reactors and they are notreliant upon the sometimes unpredictable sludge settling behavior (Geller and Gottsching,1985, Rosen and Gunnarsson, 1986, Parkin et at., 1991).The existing design and operating experience of commercially available hybrid typesystems is very limited (Beak Consultants, 1986). While the capital cost of inert supportsis a central consideration, adjuncts which substantially improve the utilization of reactorvolume have a place in current system specifications.Control of Biomass AccumulationsThe high microbial growth rates of the acidogenic microorganisms foster theaccumulation of high levels of biomass without the need of either mobile or fixed supportmedia. The upflow anaerobic sludge bed, augmented by low rates of either mechanicalstirring or liquid recycle for the purposes of pH and temperature control and for thewasting of excess biomass, is the reactor design of choice for the acid phase reactor. Themethanogens on the other hand, exhibit much lower growth rates than the acid formingbacteria or the SRB. Retention of the methanogenic bacteria in the form of granulesand/or bioflims is not dominated by concerns of excess accumulation due to the lowmethanogenic bacterial cell yield.Over an extended period of operation, unchecked biomass accumulation can leadto flow channeling, by-passing, plugging and dead volume, leading to large changes in theinternal fluid flow. This negatively affects the reactor efficiency by decreasing the contactof the effluent with active biomass. This effect is more acute at high influent wastewaterconcentrations of degradable organics for the high yield SRB and non-SRB acid producingbacteria than for the methanogens.The control of the biomass density is essential to long term system stability.Biomass concentrations in suspended growth systems can be controlled by the installation37of solid/liquid separation units such as membranes or settling columns or by a periodicwasting of sludge. The control of film thickness and of its distribution throughout thereactor is more difficult. Provided that the void spaces are not entirely occluded bybiomass, sloughing and wasting of excess biomass may be promoted by elevated rates ofgas or liquid recirculation. -MixingSome of the differences in performance of anaerobic treatment systems have beenfound by van den Berg et al. (1985) to be correlated with the liquid distribution andmixing patterns within the reactor. Mixing characteristics influence the biochemicalconversion efficiency in the liquor by: governing the contact of the microbes withsubstrate, by controlling the extent of short circuiting of substrate, and by promoting therelease of methane, carbon dioxide, hydrogen and hydrogen sulphide gases from thedigester liquor to the head space.A completely mixed reactor is relatively simple to monitor and control. However,completely mixed units demonstrate lower conversion efficiencies compared to similarunits operated in some approximation to a plug flow regime (Wilson, 1981, Li et al.,1984). As the rate of mixing is increased to approximate a CSTR (continuously stirredtank reactor), the incoming feed concentration is diluted with the reactor liquorconcentration. This has the effect of lowering the driving force of the reaction.The impact of a toxicant is a function of concentration and contact time (Parkinand Speece, 1984). Diluting an inhibitory compound to levels below its thresholdconcentration may provide for conditions which are conducive to acclimation. A CSTRprovides for maximum dilution of the in-coming feed with the reactor liquor. Thus, foreffluents which chronically contain toxicants, complete mix systems would help tominimize the inhibitory effects. While the CSTR dilutes a toxicant, it does so at theexpense of elution time. The residence time distribution of the liquid is broadened withcompletely mixed systems compared to the ideal plug flow reactor. A plug flow regime38does not dilute the incoming toxicant but it can quickly elute toxic slugs. Thischaracteristic is of benefit to limit the exposure time of sporadic elevated concentrations ofinhibitory compounds.Parkin and Speece (1984) recognized these contradictory needs of biomassretention, yet quick-elution and maximum dilution of inhibitory compounds. Theyrecommended an immobilized growth system with an optional recycle in order to meetthese needs although stirring the culture or pumping it for liquid recycle, even at low rates,is detrimental to floe formation and therefore to retention of suspended biomass.For those applications where some sort of mixing is desirable, gas mixing ratherthan mechanical mixing or liquid recycle could be considered. The most direct means ofensuring that gas mixing plays an appropriate role is to select an aspect ratio(length:diameter) to control the gas velocity of the biogas formed within the system. Thehigher the aspect ratio, the greater the gas velocity (until the gas bubbles reach a constantterminal velocity) and the mixing effect for a given wastewater treatment application.Restrictions on reactor height are strongly influenced by the form of the suspendedbacteria. While flocculated forms of bacteria can tolerate superficial liquid velocities of 0.5to 1.5 rn/hr at organic space loads of 5 to 6 kg COD/m3d, granular sludge can withstandsuperficial velocities exceeding 10 rn/hr (Henze and Harremoes, 1983).Two phase reactor systems present additional choices with respect to gas mixing.By introducing the hydrogen and carbon dioxide gas which was evolved in the acid phasereactor into the methane phase reactor, the dual effect of gas mixing and substrate additioncan have an impact on system performance. Under conditions of excessive mixing due tolarge gas production rates, the acid phase reactor gas could be diverted away from thebottom of the methane phase reactor.39Nutrient RequirementsIn order for bacterial growth and metabolic ffinctions to proceed, a number ofnutrients must be biologically available. The deficiency of any single nutrient can be ratelimiting. According to Mitchell (1974), microbial cells contain carbon, nitrogen,phosphorus and sulphur in the ratio of approximately 100:10:1:1. The commonly acceptedpractice for macro-nutrient addition is to adjust the COD:N:P ratios by adding nitrogenand phosphorus in the ratio of 350:5:1.The macro-nutrient profile also needs to be complimented by micro-nutrients inorder to achieve growth of the methanogens. Cations play a nutritional role in themetabolism of all organisms. At levels in the order of one tenth to one half of theirinhibiting concentrations, cations such as sodium, potassium, ammonium, calcium andmagnesium stimulate microbial metabolism. Many metal ions such as iron, nickel, cobaltand molybdenum are also required by the MPB in pure culture for growth and methaneformation (Archer, 1983).Since anaerobic treatment can proceed in the absence of significant growth,forecasting the required nutrient additions from the difference of the wastewatercomposition before and after treatment or from an elemental analysis of the microbes canbe expected to be unreliable. A projection of the nutrient requirements from an elementalanalysis of microbial cells gives no indication of the metabolic requirement of a particularnutrient since accumulation or adsorption onto the biomass is known to occur. Instead, amore pragmatic course is generally followed. Nutrients are added to the system such thatthey stimulate biochemical conversion rates while neither they nor their metabolites arepresent in significant concentrations in the effluent stream. Prong and Chmelauskas (1988)observed that from the addition of aluminum, calcium, cobalt, iron, molybdenum andnickel (based on an analysis of the reactor biomass) to an anaerobic UASB system fedneutral sulphite semi-chemical (NSSC) effluent that only cobalt, molybdenum and nickelwere found to be beneficial.40Reactor pHThe pH of the reactor is a result of the interaction of acidification of carbohydrateswith the carbon dioxide, suiphide and ammonia process intermediates. The reactor pH is apowerful selective agent in determining which bacterial species predominate from a mixedculture innoculum. However, the consumption -of buffering chemical for pH control is themain operating cost for the anaerobic treatment of many industrial effluents (Prong andChmelauskas, 1988). Consequently, the optimal selection of the reactor pH to maximizebiochemical conversions at minimum cost is essential to the establishment of a viabletreatment system.The non-methanogenic bacteria are not as sensitive as the methane formers andcan function over a pH range of 4.0 to 8.5. Quite generally, as the reactor pH increases tonear neutral values, the production of VFAs increases (Bull et at., 1984, Breure et at.,1986). As the reactor pH increases from 4.5 to neutral, the dominant acid phase reactorproduct shifts from butyric acid to lactic acid to acetic acid and ethanol and formic acid(Ghosh et at., 1975, Wiegant et at., 1986). Below pH 4.5, the conversion rate ofcarbohydrates decreases considerably (Hanaki et at., 1987, Kisaalita et at., 1987,Yamaguchi et at., 1989), although carbon dioxide, hydrogen, ethanol, acetate and butyrateare still produced (Hall and Adams, 1986). Above pH 4.5, the acid phase reactor productsinclude less carbon dioxide, hydrogen, ethanol and increasing concentrations of acetateand lactate. Butyric acid production predominates until pH 6.0 (Pipyn and Verstraete,1981, Wiegant et at., 1986), after which heterolactic fermentations take over as the pHapproaches neutrality and acetate and propionate are the predominant metabolites(Wiegant et at., 1986).A number of investigators (Pipyn and Verstraete, 1981, Kisaalita et at., 1987) haveconcluded from their acidogenic studies of glucose-based artificial substrates thatoperation in the pH 5,7 to 6.0 range offers the advantage of high and stable butyric acidproduction rates. Since the conversion of butyric acid to acetic acid, the primary substrate41of the methane forming bacteria, is very rapid and evolves hydrogen gas, thereforebutyrate formation is regarded as nearly as favourable as acetate.A plot of the yield of anaerobic acidogens versus pH displays a maxima from pH5.0 to 5.2 (Kisaalita et al., 1987) but there is no evidence to verifS’ the coincidence of amaximum growth rate with optimal -acidogenesis. In fact, growing cells rather thanproducing acids is a diversion of substrate carbon from methane production and anaggravation to the problem of biomass control.The trade-off in running the acid phase reactor at a lower pH and saving inbuffering chemicals lies in a depression of the rate of acidogenesis. Un-dissociated volatileacids predominate at low pH levels and these un-dissociated acids are presumed topenetrate the cell membrane more easily than the ionized form of these acids (Neal et al.,1965). Once inside the cell, the acids decrease the cell pH and consequently inhibit thebacterial metabolism. Hanaki et al. (1987) found that 95% carbohydrate degradationrequired hydraulic detention times of three times longer when operating at pH 4.5 ascompared to pH 6.0. Whether or not the trade-off is a wise one depends upon the costs ofoperating a larger reactor versus the total costs of increased chemical addition.Buffering chemicals such as sodium hydroxide, potassium hydroxide, sodiumcarbonate or sodium bicarbonate are costly and contain cations which exert toxic effectseven at fairly low concentrations. As well, the volume of buffering solution added to thesystem dilutes the concentration of substrate in the reactor and adds to the volume ofliquid for ultimate disposal.Optimum pH ranges of 6.4 to 7.6 have been reported for the methanogens (Price,1985). pH values below 6.0 and above 8.0 are very restrictive to the MPB metabolismalthough Oleszkiewicz and Hilton (1988) operated alkaline reactors up to pH 8.4 in orderto minimize the un-ionized hydrogen suiphide toxicity and recorded satisfactory levels ofmethane formation.42The sulphur reducers are very similar to the methane forming bacteria with respectto pH. The SRB pH optimum is generally accepted to be 6.8 to 7.5 (Hilton andOleszkiewicz, 1987). This optimal range appears to be species specific. For example, anumber of investigators suggested that a pH below neutral inhibited the sulphur reducingcapacity of a strain of Desulfovibrio (Leban et al., 1966, Cappenberg, 1974, Minami et at.,1988), and that growth was accelerated in the pH 7.0 to 7.5 range (Minami et at., 1988)or even 7.6 to 8.6 (Leban et at., 1966). Contrary to these observations, Badziong andThauer (1978) found that the common alkaline pH media used for these Desulfovibriocultures was less than optimal for bacterial growth and that the highest growth ratesobserved were in the pH 6.5 to 6.8 range.3.6 Two Phase Anaerobic TreatmentConventional single phase anaerobic digester systems provide one environment forvery different groups of microorganisms. However, the acid forming bacteria, the sulphurreducing bacteria and the methanogens are quite different with respect to their growthkinetics, nutritional requirements, physiology, pH optima and their ability to tolerateenvironmental stress factors such as oxygen or hydrogen suiphide (Cohen, 1983). Due tothese differences, separating the bacterial groups into VFA production and VFAconversion phases and selecting the optimal environmental conditions for both of thesegroups may be a reasonable approach towards improving the efficacy of anaerobictreatment systems.The means to achieve phase separation exploits the vastly different growth ratesand pH optima of the two bacterial groups. The SRB and the non-SRB acidogensmetabolize organic substrate faster than the MPB. From an examination of the literature,the extent of their edge in competition over the MPB for substrate would appear to be dueto a number of parameters including: microbial species, pH, the nature and concentration43of both organic and oxidized sulphur species substrate, the form of retained biomass andtemperature.Low retention times in the acid phase reactor ensure that the slow growingmethane forming bacteria are not retained to any significant extent. This favouring of thefast growing acidogenic bacteria over the methane formers in the acid phase reactor isaided by controlling the reactor pH at a level considerably lower than the 6.8 to 7.4optima of the methanogens.Phase separation is not necessarily a total elimination of all of the methane formingbacteria from the acid phase reactor or conversely the acid formers from the methanephase reactor. The actual extent of separation depends on the form in which the biomass isretained within the reactor (as biofilms or as granules), on the flow regime and onenvironmental parameters such as the dissolved oxygen content in the feed, the pH and thesubstrate available for growth.A conventional single phase anaerobic vessel cannot be equated to the sum of theacidogenic and methanogenic phases of a two phase configuration. Rather, differentbacterial populations, environmental conditions and accumulations of intermediateproducts may lead to newly favoured metabolic pathways (Barford et a!., 1986, Wiegantet a!., 1986). For example, in the acid phase reactor, the absence of methane formingbacteria can lead to an accumulation of hydrogen, and, as a direct consequence, inhibit thebreakdown of longer chained fatty acids (Archer, 1983).The overall performance of the two phase system is affected by the relativevolumes of the acid phase and methane phase reactors. The minimum generation times ofthe acid forming and methane forming bacteria are very different. Acid forming bacteriagrow approximately an order of magnitude (Andrews and Pearson, 1965) faster than themethane formers. Equal volume reactors would either dramatically underutilize the acidphase reactor volume or overload the methane phase reactor. The optimal reactor44volumes, resulting in the smallest reactors and hence the lowest capital cost, housing thesebacterial groups could not be expected to be equal.The selection of an appropriate acid phase:methane phase reactor volume ratio isinfluenced by a number of factors: the wastewater composition, the bacterial density,activity and means of retention, the environmental factors such as temperature,- pH,. .and -mixing which influence the biological growth and activity, and the variability of thewastewater hydraulic and organic loading rates. With such a large number of variables, theconclusions in the literature regarding an appropriate acid phase reactor:methane phasereactor volume ratio range from 1:1 (Lettinga and Huishof Pol, 1986) to 1:6 (de Haast etat., 1985).Advantages of Two Phase Systems1) Two phase operation affords the opportunity to create more optimal conditions for thetwo broadly defined groups of bacteria. Wasting excess biomass as well as control of pHand substrate, each at the optimal levels for the acid producing and acid consumingmicrobes, can help to increase the specific activities of the anaerobic bacterial groups(Ghosh et at., 1975, Callander and Barford, 1983a, Dohanyos et a!., 1985, Rekunen et at.,1985, Hall et at., 1987, Rintala and Vuoriranta, 1988). This promotes stability andrecovery post upset (Harper and Blaisdell, 1971, Callander and Barford, 1983b, Li et a!.,1984, Eekhaut and Alaerts, 1986, Koster, 1987, Lehtomaki, 1988, Rintala and Vuoriranta,1988).According to some of the finest researchers in the area of anaerobic wastewatertreatment, the fatty acid degradation rates in the methane reactor itself are enhanced withtwo phase operation, with findings in the literature ranging from approximately twice (Bulleta!., 1984) to three times (Cohen et a!., 1980 and Callander and Barford, 1983) that of asingle phase reactor.In spite of the claim of Beak Consultants (1986) that “There is little hard evidencethat a separate hydrolysis step has proven to be important in treating pulp and paper45effluents”, they specified a 12 hour agitated equalization basin for the Quesnel River PulpCo. BCTMP/TMP effluent treatment plant in order to control toxicity posed by peroxideand by sulphite. Vendors such as BIOTHANE, SCA and BIOTIM also include a pretreatment reactor as part of their standard designs. Acknowledged as such or not, twophase anaerobic treatment systems can be particularly useful in the wastewater—treatmenttrains of the pulp and paper industry.2) It is possible to avoid external mixing of the methane phase reactor altogether and thuspreserve the desirable approximation to plug flow kinetics with low effluentconcentrations by means of feed forward pH control. (Feedback control of methanogenicreactors is difficult due to their generally poorly mixed nature.) The target pH of themixed acid phase reactor can be set so as to result in the 6.8 to 7.4 optimum range for thesubsequent methanogenic stage. This acid phase reactor pH level can range from 4.5 to7.5, a broader range than is suitable for the methanogenic reactor. A lower acid phasereactor pH may result in lower consumption of pH buffering solution than would berequired for a single vessel operation (Hall et at., 1986).3) While the most basic function of an acid phase reactor is to hydrolyze the complexorganics into volatile fatty acids, it can also function as a surge tank and thus dampenfluctuations in incoming temperature, pH, wastewater composition and organic andhydraulic loads. This vessel is well mixed with a hardy and fast growing acetogenicbacterial population. These bacteria are not sensitive to small oxygen concentrations,environmental changes or pH (Geller and Gottsching, 1985) and are thus well suited toprotecting the more fragile methane forming bacteria.4) Through judicious selection of the reactor volumes, the organic loading rate can be setin each reactor to enhance the treatment efficiency and increase methane production rates(Andrews and Pearson, 1965, Ghosh et al., 1975, Eastman and Ferguson, 1981, Li et at.,1984, Wildenauer and Winter, 1985, Kisaalita et al., 1987). This can lead to a substantialreduction in total reactor volume or an improvement in the effluent quality for systems of46equal volume (Lehtomaki, 1988). The savings of both capital and operating costs canmore than offset the added costs for controls and instrumentation that come withincreased complexity of the physical plant for some applications (Henze and Harremoes,1983).5) A two phase process can be used to advantage in the treatment of-wastewater whichcontains compounds toxic to the methane forming bacteria but which may be degraded inan acidogenic phase (Welander and Andersson, 1985, Turk, 1988). Two examples are ofparticular relevance to this work and are worth noting:i) The acid fermenter acts to protect the slowly adapting obligate anaerobic microbialpopulation of the methane phase reactor by removing small amounts of dissolved oxygenin the feedstock (Lettinga and Hulshof Pol, 1986). Small concentrations of dissolvedoxygen can be expected to be present in the wastewater primarily due to aeration viaturbulence or mixing or passage through a dissolved air flotation clarifier.Another source of dissolved oxygen, important in some pulp and paper effluentapplications, is through the degradation of hydrogen peroxide. Welander (1988) describeda two phase anaerobic treatment system where the upstream reactor degraded peroxideand removed dissolved oxygen to protect the methane forming bacteria from upset.Dissolved oxygen is tolerated in a two phase anaerobic treatment system because,although the methanogens are strict anaerobes, some of the SRB and non-SRB acidforming bacteria can be described as facultative anaerobes.ii) Since the MPB exhibit a far greater sensitivity to sulphide levels than do the SRB or thenon-SRB acidogens (Isa et al., 1986, Oleszkiewicz and Hilton, 1986, Love, 1987,Mehrotra et a!., 1987, Hilton and Oleszkiewicz, 1988b, Maillacheruvu et al., 1993), aphysical separation of the acid producers from the methanogens may support greatertreatment efficiencies. Under high sulphur concentrations in the wastewater, the acid phasereactor is also the site of reducing the oxidized inorganic sulphur compounds to suiphide.47Acidification of the wastewater provides the sulphur reducing bacteria with hydrogen gasand VFAs which are suitable energy sources (Jensen et at., 1988).In a process train which was suggested by Mundrack and Kunst (1982) and Lee eta!. (1989) among others, suiphide would be allowed to form in the acid phase reactor andthen be removed prior to the introduction of the wastewater to the methane phase reactor.Thus, the substrate requirements could be first met for the SRB with the remainingsubstrate available for the methanogens (Potapenko et at., 1993). Puhakka et a! (1985)advocated the use of a two phase system especially for dithionate-using processes due toits demonstrated high toxicity to the MPB.A two phase anaerobic treatment system could not only avoid the suiphide-inducedinhibition of the MPB which could occur in a single phase anaerobic operation, but couldalso facilitate the delivery of the trace metal micro-nutrients to the methanogens. Tracemetals such as iron, cobalt, and nickel are precipitated as insoluble suiphides in thepresence of hydrogen sulphide and are therefore biologically unavailable. Removal ofsuiphide from the wastewater prior to adding nutrients to the methane phase reactor mayenhance the bioavailability of the metals.6) In contrast to the methane forming bacteria, the acidogenic sludge grows (and dies)very rapidly (Gujer and Zehnder, 1983). Consequently, running a settler between theacidogenic and methanogenic phases or periodic sludge wasting in order to avoid shortcircuiting, plugging or excessive pressure drops can be restricted to the acid phase reactorwithout significant loss of the methanogen bacteria. Thus, a higher proportion of activemethane forming bacteria can be retained in the methane phase reactor to result in greaterloading capacities.Disadvantages of Two Phase Operation1) The most obvious disadvantage of two phase anaerobic treatment is the increasednumber of reactor vessels required along with a corresponding increase in the need forinstrumentation and control. Thus the capital costs for a two phase treatment system may48be expected to be greater than for a single stage unit. However, since the metabolic ratesof the bacterial groups are enhanced through the appropriate selection of the reactorconditions, the required total reactor volume may be less than for single phaseconfigurations.2) The requirement for acid phase reactor pH control is the significant physical differencebetween one and two phase systems. When acid production is separated from acidconsumption, the need for pH control becomes apparent. While bicarbonate which isproduced in the methane phase reactor (MPR) is released in the effluent, a source ofalkalinity is required to counter the acidification in the acid phase reactor (APR). Theaddition of alkaline solutions such as sodium hydroxide or sodium carbonate to the acidphase reactor necessitates mixing of the reactor contents in order to disperse the chemicaland thus prevent localized toxic levels.3) Bacterial separation may exert a detrimental effect on biomass retention by changingthe flocculation and granulation tendencies and thus the settling characteristics of themethanogenic sludge. Conversely, the exclusion of the lower density and filamentous acidproducing bacteria from the methane formers may assist in the maintenance of the highdensity granules of superior settling properties. The ability to colonize inert surfaces asbioflims may also be detracted by the absence of the rapidly growing acidogens.4) Hydrogen management may present a problem upon phase separation (Archer, 1983).Single stage digesters operate in an equilibrium of hydrogen and acid production andhydrogen and acid removal. In two phase systems, unless the gas produced fromacidogenesis is transferred to the methane phase reactor, methane production willdecrease. Also, and perhaps more importantly, for effluents which contain highconcentrations of readily hydrolysable substrates, hydrogen production without removalwill lead to its accumulation in the acid reactor liquor. Elevated hydrogen concentrationshave been shown to influence the VFA distribution, shifting it away from acetic acidtowards the longer chained VFAs as a direct result of feedback inhibition of acetogenesis49(Mclnerney and Bryant, 1981). However, for sulphur rich effluents, hydrogenmanagement in the acid phase reactor should pose no difficulty since the SRB efficientlyscavenge available hydrogen. This is discussed in Chapter 4.5) An additional drawback to two phase anaerobic systems is the dearth of reportedstudies, necessitating extensive laboratory and pilot scale testing prior to implementation.3.7 Reactor TemperatureTemperature is one of the most critical environmental conditions affecting themetabolic rate of microorganisms. The temperature response of a biological process isaffected by substrate composition, concentration and the predominant population ofmicroorganisms (Lettinga, 1979).Anaerobic digestion can be conducted at the 20 to 40 °C mesophilic range and atthe 40 to 75 °C thermophilic temperature range. While the mesophilic optima ofapproximately 35 °C is generally agreed upon, the temperature optimum for thermophilictreatment is not well defined. Malina (1964), Pohiand and Bloodgood (1963) and Buhrand Andrews (1977) showed a non-continuous increase in the metabolic rate of ananaerobic digester at temperatures between 35 and 65 °C. Above the local maxima of 30to 37 °C and 50 to 60 °C, the growth rate falls off rapidly. Frostell, 1985 concluded thatthe temperature data supports 55 to 60 °C as the optimum thermophilic anaerobicdigestion temperature but McKinney (1984) found that the metabolic rates at 60 and 65°C were similar to the 55 °C optimum and that digestion at 70 and 75 °C also proceededsatisfactorily. Maatta (1985) however found that the methane production rate wasindependent of temperature from 50 to 70 °C.The optimum media temperature for the SRB, like the pH, appears to be strainspecific. While the SRB which are isolated from brackish water and marine mud samplesgrow best in the 25 to 36 °C range (Battersby, 1988) with an upper limit of about 40 °C,the same authors confess that little is known about the temperature response of non50lactate utilizing SRB such as Desuifobacter species. The temperature optima of thethermophilic SRB such as Methanobacterium thermoautotrophicum and Methanosarcinaspecies also appears to be uncertain. In practice, given the comparatively robust nature ofthe SRB, the temperature optima of the MPB will dictate the treatment system operation.The warm (50 to 80 °C) and concentrated effluents which are characteristic of therelatively closed water circuits of CTMP processes are candidates for anaerobic treatmentat thermophilic temperatures. This is shown by the thermophilic anaerobic treatment ofpulp and paper mill effluents reported by Minami et a!. (1991a), Rintala et a!. (1991),Rintala and Lepisto (1992) and Rintala and Lepisto (1993).Advantages of Thermophilic OperationThere are two important potential advantages that operation in the thermophilicrange has over conventional mesophilic operation.1) Decreased need for cooling of high temperature effluents:If it can be demonstrated that the bacteria can function effectively at highertemperatures, it may be possible to eliminate the need for operating a cooling tower andalso to avoid the accompanying problems of volatile emissions and fouling of heat transfersurfaces during the course of cooling. However, anaerobic treatment systems are generallyfollowed by aerobic polishing stages which operate at mesophiuic temperatures. Thereforethe question becomes not if; but when to cool the effluents to moderate temperatures.2) Increased rates of reaction:It is well established that, within certain limits, bacterial metabolic rates increasewith temperature. In these ranges, the growth rate is approximately linear with thereciprocal of the absolute temperature, described by the Arrhenius relationship. Thespecific growth rates of the acetate consuming methanogen Methanosarcina (Zinder andMah, 1979) and the glucose acidifying bacteria (Zoetemeyer et a!., 1982) were found tobe 40% greater at 55 °C than at 35 °C. Zinder (1988) cited the growth rate of thermophilicbacteria to be double to triple that of the mesophiles.51Bryant et a!. (1976), Varel et a!. (1977), Jewell et a!. (1978), Zeikus (1980),Schraa and Jewell (1982), Zinder (1986) and Lepisto and Rintala (1993) all observedhigher loading rates and greater stability with thermophilic compared to mesophilicoperation. Theoretically, increased thermophilic reaction rates over those of mesophilicoperation would permit the use of lower retention times at elevated temperatures and thussmaller and cheaper holding vessels could be used. Smart and Boyko (1977) suggestedthat thermophilic operations could be used in place of overloaded mesophilic units.Not all investigators have found enhanced anaerobic reactor performance atthermophilic temperatures. Whereas Rintala et at. (1991) observed rapid thermophilicsulphate reduction even with a mesophilic innoculum, Stander and Elsworth (1950) foundlower treatment efficiencies for thermophilic treatment of high sulphate wastes. van Velsenet a!. (1979) found a 25% decrease in methane production at thermophilic as comparedwith mesophilic operation. Kennedy and van den Berg (1982a) reported little differencebetween mesophilic and thermophilic operation in a DSFF reactor. van den Berg et a!.(1985) however found thermophilic loading rates and performance to be similar tomesophilic temperatures and Puhakka et at. (1988) also reported no advantage indigesting a mixture of primary and secondary sludge from a CTMP mill at 55 °C comparedto 35 °C. They demonstrated lower VS S reductions, lower tolerance to organic loadingrates, a poorer quality supernatant and significantly lower specific gas production atthermophilic temperatures.Investigations into thermophilic anaerobic treatment of pulp mill effluents havealso met with mixed results (Lee et a!. 1979). Using pilot scale studies, Salkinoja-Salonenet a!. (1983) demonstrated that 60 °C anaerobic treatment was superior to 35 °C treatmentof a TMP mill effluent. Endo and Toya (1985) found 50 °C to be the most suitabletemperature for methane fermentation of kraft effluent and Maatta (1985) found that 50 to70 °C operation proceeded at a rate 25 to 50% higher than mesophilic treatment of thesame pulp effluent. Rintala et a!. (1991) observed that a TMP effluent was anaerobically52degradable in a UASB reactor at lower retention times at thermophilic temperatures. Bothmesophilic and thermophilic operations resulted in COD removals of 65 to 70%.Thermophilic sulphate reduction appeared to have contributed to a high percentage ofCOD removal. Weigant et at. (1986) recorded the COD removal rate to be higher atthermophilic temperatures than for mesophilic systems.Disadvantages of Thermophilic OperationThermophilic anaerobic digestion also has a number of debated disadvantages.They are listed below.1. Requirements for process control:In contrast to the robust response of the anaerobic mesophiles to short termtemperature excursions, poor thermophilic process stability has been observed by someinvestigators (Garber et at., 1975, Buhr and Andrews, 1977, Lin et at., 1987).This hasbeen attributed to the reduced microbial species diversity at 55 °C. A number of reportsleads one to conclude that anaerobic waste treatment microorganisms can fianction over awide range of temperatures: from as low as 4 °C to 60 °C, as long as the temperature isstable.Mosey et at. (1971) recommended that temperatures not vary more than 2 C°/dayand temperature increments must be done slowly, at approximately 1 C°/day (Maatta,1985). Disagreeing with the conventional approach, Schraa and Jewell (1984) increasedthe reactor temperature directly from mesophilic to thermophilic settings with positiveresults whereas a gradual temperature increase was advocated by Buhr and Andrews(1977). Clausen and Shah (1981) cautioned that the acclimation of domestic sludge to thethermophilic range presented difficulties owing to the necessity of passing through the 40to 50 °C transition range. Aitken and Mullennix (1992) theorized that the differences insensitivity may be due to the way in which the microbial population was developed. Theypostulated that a significant portion of the thermotolerant organisms (that is, themesophiles which survive at elevated temperatures) would remain in the digester53population if slow rather than rapid temperature increases were employed duringthermophilic start up and that these thermotolerant mesophiles would be more sensitive tosudden temperature fluctuations.Zinder (1988) noted that the activity of the thermophilic sludge decreased withdecreasing temperature but it was still• comparable at 40 to 45 °C with that of mesophilicsludge. Smart and Boyo (1977), Varel et al. (1977) and Lettinga et al. (1991) observedthe thermophilic process to be more stable than the mesophilic operation with respect totolerating shock loads and high loads. Duff and Kennedy (1982) found just the opposite,that the thennophilic system was less able to tolerate shock loads which were routinelyhandled by the mesophilic units. The experiments of Wiegant et al. (1986) demonstratedthat short term temperature decreases had little effect but longer temperature excursionscaused a greater disturbance on the balance between the robust acetogens and the moresensitive MPB. Aitken and Mullennix (1992) observed that even room temperaturestorage for up to three months had no effect on the viability of the thermophiles.The decreased solubility of carbon dioxide with increased temperature results in adecrease in gas quality as well as significantly reduced buffering capacity (Maly andFadrus, 1971, Lin et a!., 1987). This increases the requirements for caustic solutions toregulate the reactor pH and may be a source of instability. In addition, the gas containsconsiderably more water vapour at higher temperatures, requiring design modifications inorder to dry the gas.2. Energy Requirements:If the feed stream is not released to the treatment plant already at thermophilictemperatures, the costs for heating the reactor would be excessive in order to raise thefeed temperature and to satisf,’ heat losses through digester walls (Buhr and Andrews,1977). As well, transferring heat to the effluent would be fraught with fouling difficultieson exchangers and reactor walls.543. Supernatant quality:Higher concentrations of VFAs have been observed in thermophilic reactors (vanVelsen et al., 1979, Puhakka et a!. 1988, Garber et at., 1975, Pohiand and Bloodgood,1963) compared with mesophilic runs. This appears to be due to an increased rate ofhydrolysis not in step with methane formation.4. Thermophilic seed sludge:Thermophilic anaerobic biomass is generally not available in quantities adequate tostart full scale reactors. Instead, mesophilic sludge is used and it requires an acclimation tothermophilic temperatures. Many investigators have found that mesophilic sludge is anadequate source for the start up of a thermophilic digestion system. The lower net yield asdescribed by Zeikus (1980), Henze and Harremoes (1983) and Zinder (1986) resulted in aslow start-up and poor adaptation to changes in organic and hydraulic loading rates.Therefore, biomass retention, keyed upon as the single most important variable in the highrate mesophilic designs, becomes even more essential to thermophilic operation.5. Available Experience:Laboratory, pilot or commercial scale thermophilic treatment of pulp and papereffluents have been infrequently investigated and reported. All of the full scale anaerobicsystems treating pulp mill effluents are currently operating in the mesophilic range (Lee eta!., 1989) and full scale thermophilic anaerobic treatment experience for any effluent at allis very limited. The reports which do exist are often contradictory. Perhaps this is due todifferences in the acclimation procedures, bacterial populations, effluents and theoperation of the treatment systems.The understanding of the numerous and complex metabolic pathways anaerobeswhich function at thermophilic temperatures is incomplete. The mesophilic MPB speciesand their kinetics are distinct from their thermophilic counterparts. Due to the present lackof a theoretical basis and a very limited data base, the effect of temperature on digesteroperation can only be determined experimentally.55Chapter 4The Effects of Sulphur Compounds in AnaerobicWastewater Treatment Systems4.1 OverviewSection 2 introduces the difficulties in anaerobic treatment of sulphur richeffluents.Section 3 describes the sulphur reducing bacteria.Section 4 reviews the influences that sulphate, sulphite, thiosuiphate and suiphideexert on anaerobic wastewater treatment microorganisms.Section 5 looks at the ways that sulphide impacts negatively on anaerobictreatment systems.Section 6 considers the positive effects of suiphide on anaerobic digestion.Section 7 briefly lists the options for sulphur management.4.2 IntroductionHigh concentrations of both inorganic and organic sulphur compounds are foundin many wastewaters, including effluents arising from the pulp and paper, fermentation,edible oil, petrochemical and the mining industries, Pulp manufacture uses sulphur in avariety of forms, depending on the pulping method used, and these sulphur compounds orsome derivative of them, ultimately appear in the effluent. Sulphur compounds in CTMPeffluent are divided principally as inorganic suiphite, sulphate and suiphide and organicallybound sulphur as lignosulphonates.Under anaerobic conditions, sulphate, suiphite and thiosulphate can be reduced tosuiphide. Suiphide is produced from lignosulphonates once the organic constituent isbiologically oxidized (Puhakka et a!., 1985, Sarner et a!., 1988). According to Lee et a!.56(1979), Webb (1983) and Maillacheruvu et a!. (1993), sulphur reduction can become asignificant factor in the performance and operation of anaerobic treatment of pulp andpaper effluents. Zeikus (1977) termed suiphide inhibition as “perhaps the most dramaticand environmentally significant inorganic chemical effector of methanogenesis”.The difficulties of anaerobic treatment of such sulphur rich wastewaters include thefollowing:1) Sulphur compounds inhibit biological methane formation and thus they detract fromwastewater treatment efficiency. The oxidation state of the inorganic sulphur compoundinfluences its inhibition of the methane forming bacteria.2) The sulphur reducing bacteria consume substrate at the expense of methane formation.3) Sulphur which is dissolved in the liquid effluent is present largely as suiphide followinganaerobic treatment. Suiphide is high in oxygen demand, is toxic, emits bad odours and ishighly corrosive. Consequently, the effluent needs aerobic polishing.In order to understand and to minimize the negative impact of sulphur containingwastewater on anaerobic treatment systems, it is necessary to establish the relationshipsbetween the sulphur reducing bacteria and the wastewater characteristics and theoperating conditions of the treatment system.4.3 The Sulphur Reducing BacteriaMethane formation and sulphate reduction are the two main terminal processes inthe complete anaerobic mineralization of organic matter (Schonheit et a!., 1982). Thesulphur reducing bacteria (SRB) grow and obtain energy by reducing oxidized sulphurspecies such as sulphate, sulphite and thiosulphate to form suiphide. Widely distributed,the SRB flourish in many anaerobic environments such as freshwater and marinesediments, in the rumen of ruminants, in soils, in sewage and in anaerobic treatmentsystems. The widespread nature of SRBs has led to considerable problems of odour,toxicity, destruction of concrete sewers and the acid corrosion of metals.57In the past, the SRB have been considered to comprise a small group of highlyspecialized strictly anaerobic bacteria equipped with similar physiological and bioenergeticsystems (Postgate, 1979, Peck and LeGall, 1982). This now appears not to be the case.The current literature classifies the SRB into 7 genera with a wide diversity ofmorphological and biochemical characteristics, Some of the SRB canilinction even intheabsence of oxidized sulphur compounds (Peck and LeGall, 1982).A number of species can use a wide spectrum of substrates and grow in consortiawith other microorganisms. In an excellent review of the sulphur reducers, Battersby(1988) organized the SRB into three groups:Group 1: Desuifovibrio and Desu(fotomacutum,Group 2a: the incomplete oxidizers of organic substrate, andGroup 2b: the complete oxidizers of organic substrate.Each group will be briefly reviewed.Group 1: Desulfovibrio and DesulfotomaculumDesufovibrio (desufuricans, vutgarLs, salexigens, gigas, bacutatus, and africans)is the genus most studied in the examination of waters and wastewaters (Middleton andLawrence, 1977, Nanninga and Gottschal, 1987, Lee et at., 1989). Desufovibrio aresmall, non-sporulating, gram negative, mesophilic, motile (polar flagella), strictlyanaerobic SRB which incompletely oxidize complex substrates to produce intermediatessuch as acetate and propionate. They do not utilize acetate, propionate or butyrate aselectron donors (Phelps et at., 1985, Battersby, 1988). In a sulphate-depleted medium,Bryant et aL (1977) and Phelps et at. (1985) showed that Desz4fovibrio vulgaris canoxidize ethanol or lactate to form hydrogen and acetate. Under such conditions,methanogens act as the terminal electron acceptor for the catabolic reactions performed bythe SRB.Desufotomaculum (acetoxidans and orientis) is similar to Desuifovibrio in termsof its natural sources and substrates. It is generally found in freshwater, the rumen, feces,58thermal regions and in some spoiled foods. Desufotomaculum are sporulating, gramnegative, motile rods which reduce oxidized sulphur compounds under both mesophilicand thermophilic regimes (Battersby, 1988).Group 2a: The Incomplete Oxidizers of Organic SubstrateThe bacteria which.. incompletely oxidize organic . compounds. inc1ude the- - -Desz4fobulbus propionicus and vibrioid sapovorans groups. These microbes are motile,gram negative bacteria which produce acetate from the oxidation of even numbered fattyacids and acetate and propionate from odd number fatty acids (Battersby, 1988).Group 2b: The Complete Oxidizers of Organic SubstrateThe SRB which completely oxidize organic substrate to carbon dioxide are aheterogeneous collection. This acetoclastic group includes Desu(fobacter postgatei,Desuifonema limicola Desufonema magnum, Desu41fosarcina variabilis, andDesuifococcus multivorans. They vary with respect to cell wall stain, motility, size and thesubstrates utilized, but are unified in their ability to completely oxidize their organicsubstrates to carbon dioxide while utilizing oxidized sulphur as the terminal electronacceptors to produce hydrogen suiphide (Battersby, 1988).SRB are strict anaerobes (Nanninga and Gottschal, 1987) and consequentlyrequire a low redox potential of approximately -100 my before growth can occur(Postgate, 1984). They can however tolerate a range of +115 to -450 my (Cappenbergand Prins, 1974). The lower redox potential coincides with the operative range ofmethanogenic fermentations.The degree of oxygen tolerance is strain specific (Battersby, 1988). Oxygenprotective enzymes (Hardy and Hamilton, 1981) enable the SRB to survive in someoxygenated environments such as sea water.594.4 Sulphur FormsHydrogen suiphide is formed by sulphur reducing bacteria from the reduction ofvarious oxidized inorganic sulphur compounds under anaerobic conditions. Such sulphurcompounds may include the sulphur containing amino acids such as cystine, cysteine andmethionine or the inorganic -oxidized compounds such as sulphate, sulphite andthiosulphate.According to Virkola (1983), sulphite and sulphate are the most common sulphurcompounds which are found in pulp mill effluents. Sulphate and/or sulphite are present inmost effluents originating from acid suiphite, neutral sulphite, neutral sulphite semichemical, kraft, chemi-mechanical, and CTMP mills and where aluminum sulphate is usedas a sizing agent for paper production (Habets and Kneilissen, 1985, Lee et at., 1989). Ofcourse, the oxidation state of the sulphur compounds present in the wastewater dependsupon the pulping and bleaching methods used,The inhibition of methanogenesis strongly depends upon the nature andconcentration of the inorganic sulphur compound (Khan and Trottier, 1978, Puhakka etat., 1985). Observations of Puhakka et at. (1985), Mehrotra et at. (1987) and Lee et at.,1989 produced a ranking of the inhibiting effect of sulphur compounds on methanogenicactivity. In order of increasing inhibition they are: sulphate, thiosuiphate, sulphite andsuiphide. These compounds will be discussed below.SulphateWhile the generation of methane is hindered when other electron acceptors such asoxygen or nitrate are available, the presence of sulphate in wastewaters in particular is aproblem largely due to its microbial reduction to hydrogen sulphide. Sulphate is detectedin mechanical pulping effluents although the chemical is added as sodium suiphite. Thesulphite is oxidized in the brightening stages by hydrogen peroxide, and given the rightconditions, can also be oxidized by dissolved oxygen in the wastewater.60Sulphate is used by the SRB as an electron acceptor. Hydrogen sulphide is themajor end product and to a much smaller extent, elemental sulphur or organic sulphurcompounds may also be formed (Winfrey and Zeikus, 1977, Sarner et at., 1988). At theconcentrations which are normally encountered in industrial effluents, the generalconsensus is that there are no adverse effects due to sulphate itself (Isa et at.,- 1986,Jopson et ai, 1986, Mehrotra et at., 1987, Puhakka et al., 1990).At low concentrations of 0.6 to 0.85 mM (58 to 82 mgfl), sulphate imparts abeneficial effect on the microbial ecosystem by stimulating the SRB to consume excesshydrogen and to thus maintain a minute partial pressure of hydrogen (Bryant eta!., 1977,Khan and Trottier, 1978 and van den Berg et at., 1980). A low hydrogen concentrationexerts two main effects on the microbes:1) The breakdown of long chain VFAs to produce acetate is stimulated. Hydrogen is abyproduct of the VFA degradation reactions and the reactions are thermodynamicallyfavourable only at low hydrogen concentrations.2) The rate of acetate conversion to methane can similarly be expected to be stimulated(van den Berg et a!., 1980). The competition with the SRB for hydrogen leaves the MPBwith acetate as its principal substrate. This may have the effect of decreasing the methanecontent in the gas product since carbon dioxide will not be reduced under conditions ofhydrogen limitation.Beyond the small sulphate concentration threshold required in order to stimulatethe SRB, elevated sulphate levels become problematic. At low sulphate concentrations,the decreased methane production is due primarily to the shunting of electrons from theMPB to the SRB (Isa et at., 1986). At higher sulphate concentrations, this competitiveinhibition also occurs, but it is the hydrogen sulphide product which appears to beprimarily responsible for inhibiting the MPB (Mehrotra et a!., 1987, McFarland andJewell, 1990). As a result of this observation, many authors advocate removing hydrogensuiphide from the reactor liquor in order to obtain high sulphate reduction rates. Isa et at.61(1986) and Hilton and Oieszkiewicz (1988) for example observed the rates of substrateremoval and sulphate reduction to be inversely proportional to the total sulphurconcentration in the liquor.The extent of sulphate reduction varied with the composition of the wastewateraccording to the reports of Mulder-(1984), Hilton and Oleszkiewicz (1988), McFarlandand Jewell (1990). That the wastewater composition should influence its microbialutilization would appear to be obvious, given the restricted abilities of the SRB and theMPB in particular to metabolize substrates.The literature suggests an interactive effect of wastewater strength and sulphatelevel on sulphate reduction. This is a reflection of the requirements of sufficient organicsubstrate to support the SRB in their reduction of sulphate. When the sulphateconcentration is high relative to the concentration of biodegradable organic matter,sulphate reduction dominates methane formation. A number of authors (Kroiss andWabnegg, 1983, Mulder, 1984, Lettinga, 1986, Oleszkiewicz and Hilton, 1986 forinstance) have attempted to quantifj this effect in terms of a COD: SO4 (gig) ratio. In thework of Mulder (1984), methane fermentations at low COD:SO4failed. The same authorsfound that high COD:SO4 ratios allowed stable methane formation despite high sulphatelevels. This critical COD:SO4ratio was predicted to be below 100 by Sarner et a!. (1988),33 or less by Kroiss and Wabnegg (1983), 30 by Lettinga (1986), 12 to 15 for CTMPeffluents (Pichon eta!., 1988), 10 to 15 at neutral pH (Lee eta!., 1989), and 10 or less byLettinga (1986) at influent COD concentrations of greater than 10,000 mg/i.Given the large range of values quoted in the literature, a critical COD: SO4 ratioappears to be subject to other outside factors. Hilton and Oleszkiewicz (1987) refuted aCOD: SO4 ratio as useful since both the sulphur and COD parameters aggregate a diverserange of substrates, some more utilizable than others. Rather than simply comparing thelevels of sulphate and COD, it is the concentration of the un-ionized hydrogen suiphideand biodegradable organic substrate which impacts on the treatment system.62A tremendous range of concentrations at which sulphate affects the anaerobicwastewater treatment system is evident upon examination of the literature. Winfrey andZeikus (1977) and Kristjansson et at. (1982) found 19 mg/i of sulphate to be inhibitory tothe MPB whereas sulphate concentrations of up to 10,000 mg/i were found to exert nosignificant effect on the MPB (Minami•et•aL, 1988). Numerous authors (Kristjansson etat., 1982, Isa et a!., 1986, Lettinga, 1986, Jopson et at., 1987, Mehrotra et at., 1987,Minami et at,, 1988 and Puhakka et at., 1990) have reported inconsistent results for abroad range of intermediate sulphate concentrations. This variability can perhaps beexplained by considering the influence of a number of operational parameters: wastewatercomposition, availability of substrate, the extent of sulphate reduction, the mixing regime,the bacterial species, activity, density and form of retention, batch or continuous studies,hydraulic retention time, pH and reactor temperature.SuiphiteSodium suiphite liquor is used in the chip steaming/impregnation step in the CTMPprocess. This step is not a part of TMP pulping. TMP pulp is brightened by using theBorol sodium hydrosulphite process and this also leaves a hydrosulphite and sulphiteresidual in the effluent stream. Sulphite may or may not be present in CTIvIP effluentdepending on peroxide additions (Beak Consultants, 1986) or on the presence of dissolvedoxygen. Hydrogen peroxide or dissolved oxygen in the presence of copper or cobaltcatalysts will also oxidize sulphite to form sulphate.Suiphite affects anaerobic wastewater treatment in a manner similar to that ofsulphate. Both oxidized sulphur species are terminal electron acceptors and are reduced tohydrogen sulphide by the sulphur reducing bacteria (Mehrotra et at., 1987). Sulphite andthiosulphate both increased the lag periods and lowered the methane yield in studiesperformed by Puhakka et at. (1985).Similar to sulphate, there appears to be no consensus in the literature at just whatconcentration sulphite impacts significantly on anaerobic treatment system63microorganisms. While Puhakka et al. (1984) and Pipyn et a!. (1985) described 40 mg/i ofsuiphite to be toxic to the methanogens, 5000 mg/i of suiphite was observed by Lettinga(1986) to exert no adverse effect on the treatment system. Kristjansson et al. (1982) andMehrotra et a!. (1987) found the inhibitory effects of suiphite to be only marginal.However, Puhakka et a!; (1985) observed a gas production lag period at 250 mg S032-/land a lowered methane yield at 2500 mg S032/i. Conversely, Beak Consultants (1986)concluded that the anaerobic microbial populations normally encountered in wastewatertreatment systems had no capacity to reduce sulphite and increased sulphite concentrationsled to a decreased acetate conversion rate until complete MPB inhibition occurred at 200mg S032-/l.ThiosuiphateThiosulphate transformation under anaerobic conditions has been infrequentlyreported (Badziong and Thauer, 1978, Trudinger, 1978, Domka and Szulczynski, 1980,Postgate, 1984). The reduction of thiosulphate to suiphide with lactate as the carbonsource was proposed by Domka and Szulczynski (1980) as:2CH3CHOHCOO + S203 -* 2C113C00 + 112S + S2 +1120+ 2C0This thiosulphate reduction reaction proceeds via sulphite (Nakatsukasa and Akagi, 1969,Haschke and Campbell, 1971, Hatchikian, 1974). Consequently, the concerns of theeffects of sulphite on anaerobic treatment micro flora can also be shared for thiosulphatelevels.SuiphideSuiphide is formed by the biological reduction of inorganic sulphur compoundssuch as sulphate, thiosulphate and suiphite and of sulphur containing amino acids. In thekraft process, suiphide is formed in the pulping stage, following the addition of sulphurwhich is used to prevent the degradation of cellulose. Suiphide is biologically formed inpulp mill effluents by SRB which are attached as slimes or are present in the stagnantwater zones of the machine system (Webb, 1985).64Hydrogen suiphide is distributed betweenH2S(g),H2S(aq), HS(aq), S2(aq), andmetal sulphide precipitates. The pH governs the ionic form of hydrogen sulphide and itsdistribution between the effluent liquor and the gas phases. Hydrogen sulphide is highlysoluble and it forms a weak acid in solution. It dissociates in two steps (Callander andBarford, 1983b). -H2S HS- +HS ÷s2- +At the approximately neutral pH range of anaerobic treatment systems, the seconddissociation can be considered to be insignificant (Winfrey and Zeikus, 1977, Isa et a!.,1986).Given the pH of the mixed liquor, the concentration of the free or unionizedhydrogen suiphide in solution can be calculated. It is the product of! the decimal fractionof the aqueous sulphide concentration which is present as free hydrogen sulphide, and theconcentration of total dissolved suiphide (H2S + HS- + S2-). f is calculated using thefollowing equation (Hilton and Oleszkiewicz, 1988):f= 1/ji + 1.28 exp (pH-7)j (1)The total sulphide in the system is partitioned between the gaseous and liquidphases. In the gas phase, sulphide is in the hydrogen suiphide (H2S) form. In the liquidphase, the suiphide is divided between hydrogen sulphide and bisuiphide ion (HSj. Theequilibrium of hydrogen suiphide between the gas and liquid phases is governed by Henry’slaw:[H2S]aq = a[H2S]g (2)The absorption coefficient, a, is a function of temperature and has a value of 1.83 at 35 °C(Isa eta!., 1986) and 1.59 at 55 DC (Wilhelm et a!, 1977).Even at low concentrations, suiphide disturbs biological methane production. Dueto its most profound effect on organic removal efficiencies, the presence of hydrogen65suiphide in liquid effluents is a problem of considerable importance to anaerobic treatmentapplications both within and outside of the pulp and paper industry.4.5 Why is Sulphide a Problem?a) Inhibition of MPBElevated hydrogen suiphide concentrations impair methane production, resulting inan accumulation of VFAs, especially of acetate. However, an examination of the literaturereveals no consensus on the threshold concentration at which hydrogen sulphide inhibitsmethane formation. The inhibitory value most often quoted is 200 mg H2S/l (Mulder,1984, Olthof et a!., 1985, Orivuori, 1985, McFarland and Jewell, 1990), but this figurevaries widely. Minami et a!. (1988) found concentrations as small as 34 mg/i of suiphide toinhibit the methanogens. Even at suiphide levels up to 1000 mg/i, the investigators founddifferent responses. Hilton and Oleszkiewicz (1988) reported an uninhibited system at thisconcentration whereas Isa et a!. (1986) observed half the methanogenic activity andPuhakka et a!. (1985) and Hilton and Oleszkiewicz (1987) found complete inhibition. Atplay here are the factors of wastewater composition, availability of substrate, the extent ofsulphate reduction, the mixing regime, the bacterial species, activity, density, form ofretention and acclimation, batch or continuous studies, hydraulic retention time, reactorpH and reactor temperature.According to Schlegel, (1969) and Hilton and Oleszkiewicz (1987), it is theelevated concentrations of un-ionized suiphide that both the SRB and the MPB aresensitive to. Neale et a!. (1965) theorized that the uncharged hydrogen suiphide moleculescan pass through the cell membranes by ordinary difii.ision more readily than chargedmolecules can.The mechanism of sulphur induced inhibition of MPB is not fully understood(Winfrey and Zeikus, 1977, Ingvorsen et a!., 1984). Several mechanisms have beenproposed (Ingvorsen et a!., 1984). The MPB may be inhibited by toxic metabolites or66there may be a lack of methane precursors or a lack of required growth factors. It isspeculated that the methanogenic pathways may be blocked by the presence of suiphide orsulphur containing compounds (Cappenberg, 1974, Abram and Nedwell, 1978,Oleszkiewicz and Hilton, 1986). Another suggested mechanism is that the SRB have moreefficient uptake systems than the MPB for the substrates utilized by both groups.-— -b) Competition of SRB with MPB for SubstrateIn mixed cultures and when non-limiting quantities of sulphate are present, theSRB occupy essentially the same ecological niche as the methanogens. They both utilizethe same end products from acidogenesis, that is, the methanogenic precursors acetate andhydrogen (Middleton and Lawrence, 1977, Winfrey and Zeikus, 1977, Kristjansson et at.,1982). In addition, formate, propionate, lactate, ethanol and pyruvate are also used by theSRB (Butlin et at., 1956, Archer, 1983, Wu et a!., 1991). A number of investigators(Kristjansson et at., 1982) have concluded that the MPB are not inhibited per se by theactivity of the SRB but they rather have a lower affinity for the substrate which they canboth use. This results in suppression of the MPB. The MPB can co-exist with the SRBhowever. The outcome of the competition depends upon the sulphate availability and theresulting hydrogen suiphide concentration, the substrate mix and the relative sizes andactivities of the microbial populations (Lovley et a!., 1982).The SRB are better competitors than are the MPB for a number of reasons. Theyare less fastidious anaerobes (Hilton and Oleszkiewicz, 1987) and appear to be moreadaptable to changing reactor conditions such as pH and temperature variations (Winfreyand Zeikus, 1977, Jensen et at., 1988). The SRB are only slightly affected by hydrogensuiphide concentrations as compared to the MPB (Isa et at., 1986) and, in the presence ofa sulphur source, can easily outcompete the MPB under conditions of extremely lowcarbon concentrations (Hilton and Oleszkiewicz, 1987).The SRWs ability to outcompete the MPB for substrate is supported by thethermodynamic argument of Claypool and Kaplan (1974), Martens and Bemer (1974),67Abram and Nedwell, (1978) and Kristjansson et a!., 1982. The chemical reactions formethane formation and for sulphate reduction along with their standard free energychanges are listed below (Winfrey and Zeikus, 1977, Kristjansson et a!., 1982):Methanogenesis:CO2 + 4H2 -* CR4 + 2H0AG° = -135.1 KJCH3OOH -* CR4 + CO2AG° =-28.5KJSulphate Reduction:S042 + 4H2 —* H2S+H0+20WAG°=-154.OKJSO42 + CR3OOH -> H25 + 2HC03AG° = -47.3 KJThe SRB and MPB compete for available substrate. Per mole of reactant andunder standard conditions, the reduction of carbon dioxide by hydrogen to producemethane yields 135.1 KJ, whereas reduction of sulphate to hydrogen sulphide yields 154.01G. Acetate conversion by methanogens to methane and carbon dioxide yields just 28.5 KJcompared to a yield of 47.3 1(3 when utilized by the SRB. Therefore, under standardconditions, the SRB are energetically favoured over the MPB in the utilization of bothhydrogen and acetate. If there is a sufficient SRB population density of hydrogen andacetate users, the SRB should outcompete the MPB for both hydrogen and acetate.However, the free energy changes which are pertinent to the chemical reactions occur inliving cells are not at standard conditions.Isa et at. (1986) found that biomass retention affected the bacterial population’sability to compete. They observed that the MPB were able to outcompete the SRB in amobile fixed film reactor due to their higher propensity to form films on inert surfaces. Not68all SRB or MPB metabolize acetate however, and the preferential consumption by SRBover MPB is contingent upon the presence of acetate utilizing SRB.Thermodynamic calculations are employed to predict whether or not a reaction isfavoured. It isn’t valid however, to claim that the thermodynamically favourable reactionwill totally exclude another favourable reaction• which - yields less energy. Sincethermodynamics gives no indication of reaction rate, system performance cannot bepredicted on the basis of thermodynamics alone without kinetics.Competition may have a kinetic rather than purely a thermodynamic basis(Puhakka, 1990). The SRB in general have a lower Ks (the substrate concentration atwhich the reaction is half the maximum rate) for hydrogen than the MPB (Kristjansson etal., 1982). Recent studies on SRB/MPB coculture kinetics have demonstrated the MonodKs acetate value is at least 10 times lower for the SRB than for the MPB (Schonheit et a!.,1982). According to Veldkamp and Jannasch (1972) this difference is significant wheneven minor differences in Ks values are sufficient to enable one organism to outgrow theother in a substrate limited chemostat. The low Ks implies that the SRB shouldoutcompete the methanogens, especially at low levels of hydrogen and acetate. Accordingto Bhattacharya (1986), certain acetoclastic methanogens have a low Ks, comparable tothe Ks of the SRB which also metabolize acetate. These MPB could compete with theSRB for acetate.Methane production and sulphate reduction are not mutually exclusive. Withexcess hydrogen and providing that suiphide accumulations do not become inhibitory, theyhave little effect on each other. In experiments which were designed to demonstrate thiscompetition for hydrogen, a number of authors have documented the restoration ofmethane production by supplying excess hydrogen gas (Winfiey and Zeikus, 1977, Phelpset a!., 1985). When hydrogen is rate limiting however, competition between the MPB andthe SRB occurs. In cultures containing both MPB and SRB when sulphate concentrationsare not rate limiting, the SRB inhibit methane production by more effectively scavenging69hydrogen. This lowers the hydrogen partial pressure below a threshold level necessary forhydrogen utilization by the MPB (Abram and Nedwell, 1978). Sulphate dependentinterspecies hydrogen transfer from the MPB to the SRB results in less methaneproduction, increased carbon dioxide formation and increased sulphide formation (Winfreyand Zeikus, 1977, Phelps etal.,l985);Many investigators (Ueki et a!., 1986, Nanninga and Gottschal, 1987) haveobserved that the contribution of acetate as an electron donor for sulphate reduction inanaerobic treatment systems is very small. The SRB do oxidize VFAs such as lactate,propionate, butyrate, as well as ethanol as electron donors and produce acetate (Ueki eta!., 1986). This acetate is then available to the MPB.In the absence of oxidized sulphur species such as sulphate, a number ofinvestigators have found evidence of a commensal association between MPB and SRB(Cappenberg, 1975, Bryant et a!., 1977, Winfrey and Zeikus, 1977, Phelps et a!., 1985),although this relationship is complicated by the release of hydrogen sulphide which inhibitsthe MPB growth. These experimenters have reported the acetate and hydrogen whichwere produced by the SRB are utilized by the MPB under sulphate limited conditions.At the elevated substrate concentrations which are typical of anaerobic treatmentsystems and especially at high acetate levels, the IvIPB can reportedly outgrow the SRB(Isa et a!., 1986b, Yoda et a!., 1987). This relatively non-competitive nature of the SRBcontrasts with observations of the SRB’s dominance in nutrient poor natural environments.The SRB outcompete the ]VIPB for hydrogen and acetate at low organic loading rates orat low substrate levels, say less than 0.5 g COD/l (Isa et al., 1986, McCartney andOleszkiewicz, 1990). In experiments which were conducted by Isa et a!. (1986), loweringthe influent COD from 5.0 to 0.5 g/l enabled the SRB to divert the electron flow from 11to 34% at the cost of decreasing the specific methane yield from approximately 300 to 200ml CH4/g CODr.70c) Impact on Receiving WatersHydrogen suiphide impacts on the receiving waters both directly and indirectly.Acting directly, suiphide is harmful to aquatic life at very low levels, with 0.018 to 0.35mg sulphide/l the 96hLC50 for freshwater fish or 0.05 to 3 mg sulphide/l the 96hLC50 forsaltwater fish (Environment Canada,- 1984). The large range in these values reflects thevarying toxic impact of sulphide on the fish, dependent upon species and age of the testorganism.Impacting indirectly, hydrogen suiphide has a high oxygen demand. When onemole of hydrogen sulphide is oxidized to form one mole (96 g) of sulphate, 2 moles or 64g of oxygen are needed. Therefore, dissolved sulphides in the effluent elevate the CODand necessitate post oxidation treatment prior to water reuse within the mill or prior todischarge into receiving waters. For the sulphur rich wastewaters which are characteristicof mechanical pulping effluents, anaerobic treatment generates an exit stream with a highsulphide oxygen demand (up to 500 mg/i for the experiments conducted in this work). Thepresence of sulphide alone can rapidly diminish the dissolved oxygen levels in receivingwaters.d) Odour, Occupational Health and CorrosionHydrogen suiphide is renown for its unpleasant rotten egg odour. The odourthreshold for hydrogen sulphide ranges from 0.000 1 to 0.00 1 ppm (Environment Canada,1984). Although the characteristic rotten egg odour provides a warning of its presence,the individual’s sense of smell is quickly fatigued by hydrogen suiphide.Hydrogen suiphide is highly toxic to people by inhalation and by contact, acting asa respiratory irritant at concentrations of 50 to 500 ppm, and, at elevated concentrations,as a systemic poison, causing unconsciousness and death (ACGIIT, 1988). The thresholdlimit value (TLV) is 10 ppm and the short term exposure limit (STEL, 15 minutes) is 15ppm (ACGIII, 1988).71Hydrogen suiphide is notorious for its corrosive properties when in the presence ofwater. Boilers, internal combustion engines, concrete surfaces and steel pipelines requireprotection when exposed to hydrogen sulphide.4.6 Benefits of Hydrogen SuiphideIn spite of its negative impact on anaerobic treatment systems, hydrogen sulphidealso plays a number of beneficial roles.a) Metabolic Requirement for SuiphideThe literature on anaerobic fermentation indicates that sulphur compounds, inparticular sulphide, are at the same time both necessary yet undesirable (Zeikus, 1977,Ronnow and Gunnarsson, 1981). While elevated concentrations of sulphur compoundscan create problems, small sulphur levels are essential to the treatment system microbes.A constituent of the sulphur containing amino acids and of a number of vitaminsand coenzymes, sulphur is an essential element for the growth and activity of all livingcells (Diekert et al., 1980, Ronnow and Gunnarsson, 1981). Also, methanogens possess aunique sulphur rich cofactor known as Co-enzyme M (HS-CH2-CH03 “2-mercaptoethane sulfonic acid”) (McBride and Wolfe, 1971, Balch and Wolfe, 1976,Ronnow and Gunnarsson, 1981, Scherer and Sahm, 1981). The IvIPB also contain a largenumber of iron-sulphur proteins which are thought to take part in electron transport(Ronnow and Gunnarsson, 1981, Hilton and Oleszkiewicz, 1987).In pure culture studies, depending on the MPB species and the process conditionssuch as pH, temperature and nature of the wastewater, the beneficial sulphideconcentration is reported to range from 3.4 to 50 mg S2/l (Zehnder and Wuhrmann,1977, Mountfort and Asher, 1979, Ronnow and Gunnarsson, 1981, Callander andBarford, 1983, Nanninga and Gottschal, 1987). Such concentrations are dependent uponthe microbial consortia and the process conditions.72b) Precipitation of MetalsIn addition to satisfying a metabolic requirement, suiphide exerts an indirect effectby precipitating non-alkali metal ions from solution (Middleton and Lawrence, 1977,Mountfort and Asher, 1979, Scherer and Sahm, 1981, Isa et al., 1986). Since the bioavailability of metals depends only on their aqueous concentration (Lawrence andMcCarty, 1965), precipitation as insoluble metal sulphides can affect the bio-treatmentsystem. A number of metals such as iron, cobalt, copper, nickel and molybdenum areessential growth factors at low levels but they impart inhibitory effects at elevatedconcentrations (Masselli et a!., 1961, Lawrence and McCarty, 1965, Middleton andLawrence, 1977, Mountfort and Asher, 1979, Isa et a!., 1986). These metals areeffectively removed from the liquor due to the low solubility products of their metalsuiphides (Maree and Strydom, 1985). The earth metals, calcium and magnesium, and thealkali earth metals, sodium and potassium, and manganese have high metal suiphidesolubility products and are little influenced by suiphide (Maree and Strydom, 1985).Since one mole of suiphide is required per mole of heavy metals for precipitation(McCarty, 1964), Speece (1983) and Callander and Barford (1983) advocated slug dosingof micro-nutrients at levels exceeding those precipitated as suiphides in order to tip thebalance from metal suiphide precipitation to uptake by the microorganisms.c) Oxidation of Organic CompoundsAnother benefit of the presence of inorganic sulphur compounds in effluents is thecontribution which the SRB can make towards the oxidation of organic compounds(Hilton and Oleszkiewicz, 1987). This would appear to be due to the high requirement fororganic substrates by the SRB. The SRB utilize a number of complex VFAs and alsometabolize the hydrogen by-product. This multiple role of acetogen, acetoclast andhydrogen sink enables the SRB to contribute to both the stability and the rate of organicsremoval by the treatment system. The net impact of the SRB on the organic profile of an73effluent is to decrease the concentration of complex VFAs and render the wastewatermore treatable by either anaerobic or aerobic means.While the reduction of oxidized sulphur compounds to suiphide is coincidentalwith appreciable COD removal, this organic removal efficiency becomes progressivelyworse with increased concentrations of reduced sulphur compounds in• the effluent,inhibiting the MPB and/or the SRB (Oleszkiewicz and Hilton, 1986, McFarland andJewell, 1990).An additional benefit of sulphur reduction is a pH effect. The SRB generatealkalinity from the reduction of sulphate, 3.1 mg of calcium carbonate alkalinity per mg ofsulphate reduced (Middleton and Lawrence, 1977). This decreases the requirements forbuffering chemicals (Oleszkiewicz and Hilton, 1977).4.7 Options for Sulphur ManagementIn order to minimize the disruptive effects of sulphur compounds on the microbialcommunities which are found in anaerobic treatment systems, control of the un-dissociatedhydrogen sulphide concentration is essential. There are three principal means toaccomplish this.1) Raise the pH of the digester liquor to the alkaline range to ensure that the suiphide isionized and therefore less inhibitory to the MPB.2) Prevent the sulphur reducing bacteria from functioning altogether.3) Allow biological sulphur reduction and the remove the sulphide from solution by usingeither precipitation or gas stripping.Each of the three principal sulphur management options will be reviewed in greaterdetail in subsequent chapters.Other methods of responding to the inhibitory effects of sulphur on the treatmentsystem microorganisms include:741) Dilute the wastewater (during reactor start up in particular) with non-toxic wastewaterssuch as effluent from a downstream aerobic treatment unit. This approach to reduce theconcentrations of the offending compounds to below their threshold levels has beenadvocated by Beak Consultants (1986), Habets and de Vegt (1990) and Sierra-Alvarez eta!. (1993).2) Physical pre-treatment such as:i) oxidizing sulphite to sulphate using ozone, peroxide or air (Eeckhaut and Alaerts,1986),ii) air bubbling at pH of 2 and 70 °C (Eeckhaut and Alaerts, 1986, Gunnarsson et at.,1989), oriii) precipitating sulphate using barium hydroxide or chloride salts.Physical pre-treatment options are generally consumptive of both chemicals andenergy. Diluting the effluent to below the threshold inhibitory concentration can addsignificantly to the treatment vessel volume or increase the liquid velocity through thereactor. For the wastewater volumes and sulphur concentrations generated by pulp mills,these physical treatment methods appear to have little practical value (Maree and’Strydom, 1985, Turk, 1988, Lettinga et al., 1991).75Chapter 5Experimental Materials and Methods5.1 OverviewSection 2 describes the inocula which were used in these experiments.Section 3 describes the feedstock which was obtained from Quesnel River Pulp.Section 4 details the laboratory scale anaerobic treatment system which wasoperated for these experiments.Section 5 explains why the parameters which were used to describe theseexperiments were chosen.Section 6 describes the experimental protocol.Section 7 lists the analytical procedures used.Section 8 briefly explains the plan of experiments which was followed for thiswork.5.2 InoculaMesophiuic anaerobic sludge was shipped to the UBC Pulp and Paper Center fromthe anaerobic treatment system which was operated by the Quesnel River Pulp Co. to treatsubstantially the same effluent as was used in this study. Sludge was refrigerated at 3 °Cuntil it was needed for the experiments. The microorganisms warmed up to roomtemperature overnight prior to pouring them into the reactors for operation at 35 °C.Thermophilic experiments were performed after the mesophilic runs were complete usingthe same inocula. Temperature increments were targeted at 1 C° per day in moving fromthe 35 °C to the 55 °C operation.765.3 FeedstockQuesnel River Pulp BCTMP/TMP effluent, in a volumetric ratio of approximately2:1 BCTMP:TMP with 65% to 70% white water and 30% to 35% waste water, was usedfor this study. The white water stream is characterized by high concentrations of difficultto settle fibres and fines, soluble COD and toxicity (Rankin el at., 1992). The waste watercontains effluent discharged from the reject control, chip conditioning and chipimpregnation units. It is high in suiphite and poorly settleable solids but contributescomparatively little to either COD or toxicity (Rankin et a!., 1992). The Quesnel RiverPulp mill typically produces 300 to 450 m3/h of white water and 50 to 150 m3/h wastewater. A typical furnish for the mill was 60% white spruce and 40% lodgepole pine.Douglas fir and aspen were also used infrequently and in much lesser amounts.Effluent from Quesnel River Pulp was withdrawn directly following the KROFTADissolved Air Flotation Clarifier unit which removed, with the help of a phenolic resin andpolyethylene oxide, large quantities of fibres and fines. Effluent was transported by truckfrom the Quesnel River Pulp Co. in one cubic meter plastic totes. At the UBC Pulp andPaper Center, the effluent was pumped into 20 liter plastic pails and refrigerated at 3 °Cfor up to 2 months. Sediment and fine particulates were screened from the liquid just priorto use.Urea and phosphoric acid were added to the fresh wastewater daily in theCOD:N:P ratio of 350:5:1 in order that nitrogen and phosphorus were not limiting.Chlorides of iron, cobalt and nickel were also constituents of the concentrated stocksolution which was added to the effluent. Oleszkiewicz and Romanek (1989) observedthese cations to contribute to system performance and maintenance of the biomass ingranular form. They were added to the Quesnel River Pulp effluent to result inconcentrations of 10,1 and 1 mg/i for the iron, cobalt and nickel respectively. The additionof the stock nutrient solution diluted the effluent by just 1%.775.4 ApparatusA laboratory scale anaerobic treatment system was designed and constructed in theUBC Department of Chemical Engineering. It was comprised of three parallel dual reactorsystems and each set was a different volume. The different reactor volumes enabled theeffect of HRT to be assessed with an equal wastewater- flow rate applied to each reactorset. A schematic of one reactor set of the apparatus is given in Figure 2. A detaileddescription of the apparatus follows.A variable speed Masterfiex peristaltic pump delivered fresh wastewater from a 20liter plastic pail at ambient temperature to the base of the acid phase reactors. Thesereactors were designed as upflow anaerobic sludge bed systems, equipped with liquidrecycle, but without the granular biomass or gas separation devices typical of suchsystems. These three acid phase reactors of 2.5, 3.5 and 4.5 inch (6.4, 8.9, and 11.4 cm)inside diameters, each 12 inches (30.5 cm) tall, gave empty reactor volumes ofapproximately 1.0, 1.9 and 3.1 litres. The acid phase reactors were housed in a commonwater bath for temperature control by a Cobra bath water heater and circulator. Reactortemperatures were measured by copper constantan thermocouples and were indicated onan Omega model 650 multi-channeltemperature meter. Two Omega temperature controllers (one for the acid phase reactorbath, the other for the methane phase reactor bath) provided an upper limit to the waterbath temperature in order to protect the system against malfunction.Control of the acid phase reactor pH was achieved by recirculating the acid phasereactor liquor by means of a variable speed Masterflex peristaltic pump from the top of thereactor past a pH probe and back to the base of the reactor. A 0.5 N sodium carbonatesolution was pumped into the acid phase reactors when the setpoint of a Cole Parmerseries 7142 pH control pump system was surpassed. Float switches were installed in boththe feed tanks and in the sodium carbonate tanks in order to prevent the introduction of airshould the respective pumps exhaust the liquid contents.78To Gas Measurement and VentFigure 2: Experimental ApparatusDrainGas ScrubberGas. SampleWasteFeCI3Water TrapIn-LineOverflowTeeH20pH Probe H20Sodium Carbonate Feed Tank Weigh Scale79Effluent from the acid phase reactors flowed past 1/3 psig check valves into themethane phase reactors of the same diameter as the acid phase reactors. The check valveshelped to minimize any liquor and biomass back flow into the acid phase reactors eitherduring the course of sampling or in the event of leaks. These reactors were 36 inches (91cm) tall. The liquid level was set at approximately 24 inches (61 cm) to result in liquidvolumes of 1.9, 3.8 and 6.3 litres, double those of the acid phase reactors. The total liquidvolume of the three parallel reactor sets was 2.9, 5.7 and 9.4 litres respectively. The 12inch (30.5 cm) head space of each methane reactor was designed to accommodatefoaming as well as the minor fluctuations in liquid level which might be expected to occur.Each of the methane phase reactors held a cartridge of 1/8 inch (3 mm) thick,needle punched, polyester cloth wound on 1/16 inch diameter (1.6 mm) stainless steel rodsin the shape of a coiled hexagon with 1/2 inch (1.3 cm) spacing between layers of thespiral. These bioflim support cartridges were the length of the reactors. Kennedy et al.(1988) selected the same material because it was proven as an effective support mediumand it was easy to fabricate and to install in the laboratory scale reactors. This materialprovided a wetted surface area of 143, 462 and 836 square inches (923, 2980 and 5390cm2) respectively for the 2.9, 5.7 and 9.4 litre reactors. This resulted in material surfacearea:reactor volume ratios of 32, 52 and 57 m2/m3 and void fractions of 86, 82 and 78%respectively. Provision of an inert surface for attachment by the microbes was intended toaugment the retention of the granular and flocculant biomass originally present in the seedinocula.The three methane phase reactors were housed in a common water bath fortemperature control using a Cobra bath water heater and circulator. There was noprovision for direct pH control for these generally unmixed reactors. Rather, the methanephase reactor pH values were a response to the acid phase reactor pHs and the reactorconditions. An adjustable height effluent overflow tee controlled the liquid level at80approximately 24 inches (61 cm). Effluent was weighed and was then flushed down thedrain with tap water in order to minimize offensive odours.Both of the reactor sets were constructed of clear Plexiglas. All tubing wetted bythe liquid effluent was made of polypropylene with the exception of short lengths ofMasterfiex silicone tubing which was used in theperistaltic pump heads.- -The gas which was produced in the reactors was contacted with iron spongebundles contained in 0.68 liter plastic tubes in order to protect the wet test meters fromthe corrosive effect of hydrogen suiphide. The hydrogen sulphide in the biogas reactedwith the iron to form an insoluble iron suiphide. The gas, once measured in 5 liter perrevolution wet test meters, was vented into a flume hood.In the case of the gas scrubbing and recycle experiments to strip hydrogen sulphidefrom solution, the evolved gas was pumped from the reactor head space by a Masterflexvariable speed pump. It was then bubbled through 0.68 litre scrubbers filled with 200 g/lferric chloride (FeC13’6H20) solution to precipitate iron suiphide and then it wasreintroduced into the base of the methane phase reactors. The gas recycle flow rate wasset at an average of 200 mllmin, controlled and measured by Cole Parmer rotameters. Thepartially spent iron chloride solution was replaced daily with fresh solution. A half filledand sealed, one liter Erlenmeyer flask, water trap prevented back flow of gas through thewet test meters.The odours arising from hydrogen sulphide and other reduced sulphur specieswhich emanated from the system were withdrawn by the laboratory ventilation systemfrom the enclosed area.5.5 AnalysesGenerally, it is difficult to measure the efficacy of a treatment system by directmeans such as measuring the individual concentrations of the consortia of microorganisms.Toerien and Hattingh (1969) and Wuhrmann (1982) warned of the methodological81difficulties in counting microorganisms which possess a large range of metabolic activities.They concluded that this unreliable procedure should never be used. Labat and Garcia(1986) found that the total bacterial number had no correlation with digester performanceand was of no aid in assessing the state of an anaerobic digester.Instead of counting microorganisms, they are selected by the -plant- design-andoperating conditions and the activity in the reactor is assessed by measuring theconcentrations of products and reactants. The destruction of organic compounds and thegas produced, reactor pH, volatile fatty acids, carbohydrates and individual inorganicsulphur compounds are factors which can be measured and used as indicators of digesterperformance. The individual parameters were assayed according to the procedures asoutlined below.The analyses which were performed on the fresh feed, the feed in the tank, the acidphase reactor effluents and the methane reactor effluents were as follows.Feedstock: chemical oxygen demand, pH, sulphate, suiphite, thiosulphate,total dissolved sulphide, total carbohydrates, total organic carbon,tannin-lignin, and volatile fatty acidsReactors: gas composition: carbon dioxide, methane, hydrogen suiphide, pH,sulphate, suiphite, thiosulphate, total dissolved sulphide, totalcarbohydrates, total organic carbon, volatile fatty acids, carbondioxide gas, hydrogen sulphide gas, and methane gas5.6 Duration and Sampling of RunsExperiments were conducted to assess the quasi-steady state responses(approximately 5% coefficient of variation (CV) for sulphate, suiphide and TOC) to aseries of sulphur management techniques. After a change in sulphur management strategyor wastewater batch, at least one week or a minimum of 4 residence times was allowed foracclimation prior to any sampling. At least ten samples were taken over the next fourteen82days or until steady state was approximated. This sampling protocol allowed for at leastten complete residence times per experimental run. Longer runs were warranted wheninstability was evident as indicated by varying (standard deviations exceeding 5 % of themean) volumetric loading rates, gas production rates, pH, rate of sodium carbonateconsumption, total dissolved sulphideorsulphate concentrations.Gas production, effluent volume, pH and temperature were logged each day. Thetotal dissolved sulphide and sulphate concentrations of the fresh feed, the feed tank, thethree acid phase reactors and the three methane phase reactors were measured eachsampling day. The daily rate of consumption of 0.5 N sodium carbonate solution wascalculated from the approximately weekly additions to the sodium carbonate tanks. Liquidsamples were taken daily and were centrifuged (35000 rpm for 15 minutes) with thesupernatant frozen until convenient for further assay. Parallel tests were performed onselected fresh and frozen samples in order to ascertain the validity of this sample storagepractice.5.7 Analytical ProceduresSolidsThe standard methods commonly used for wastewater treatment analysis are oflimited benefit in evaluating the efficacy of the anaerobic digestion. Accurate measurementof solids destruction is difficult and the normal variability of the feed stock tends to masksubtle changes in reactor performance. Representative sampling of unmixed sludgebedlbiofilm reactors is prone to error, especially with respect to the solid phase.Suspended solids measurements tell very little about the form of the insoluble material.High effluent suspended solids for example could reflect either the self-regulating aspectof the reactor and its elevated biomass growth rates or low conversion of influentsuspended solids.83Analyses for total solids (TS), volatile solids (VS) and suspended solids werecarried out only to characterize the feed according to the procedures outlined in theStandard Methods for the Examination of Water and Wastewater (APHA, 1989). Volatilesuspended solids, a coarse measure of biomass, was not assayed due to the impracticalityof sampling from fixed bed reactors.Chemical Oxygen DemandThe chemical oxygen demand of the feedstock was measured in order tocharacterize the substrate and to compare this work with reports found elsewhere in theliterature. The reactor CODs were not measured due to high levels of suiphide formed.Since suiphide is high in COD and is also very labile, samples which were rich in dissolvedsuiphide would be subject to considerable error. For some of the experiments where thetotal dissolved sulphide levels exceeded 250 mg S2-Il for example, the contribution ofsuiphide to the COD would have been in excess of 500 mg/I.COD tests were performed on a Hach 2000 spectrophotometer at 480 nm andwere calibrated against prepared glucose solutions.Total Carbon and Total Organic CarbonTotal organic carbon was assayed in order to measure the wastewater treatmentefficiency. Total organic carbon measurements have the same attributes as COD but arenot interfered with by dissolved hydrogen suiphide. TOC tests were run on pooledsamples for all of the experiments.Total carbon and total organic carbon analyses were performed on two differentinstruments. The first was an Astro Model 1850 TOC-TC Analyzer, complete with anAstro 910 200-sample auto sampler. The assay employed the infrared measurement ofcarbon dioxide after the oxidation of liquid samples using sodium persuiphate, oxygen andUV light. The second instrument was a Shimadzu TOC-500 Total Organic Analyzerequipped with an ASI-502 Automatic Sample Injector. This system used a total carboncatalyst and a non-dispersive infrared gas analyzer.84Total organic carbon differs from total carbon in that TOC does not include carbonfrom dissolved carbonates. Dissolved carbonates were removed from solution byacidification with concentrated phosphoric acid to a pH of less than 2.0, followed bysparging with nitrogen gas for several minutes in order to remove any carbon dioxide. Thisremoval of the dissolved carbonate was an automated feature of the carbon analyzer. Thesystem was calibrated against a urea solution standard and a sodium carbonate andbicarbonate solution for total carbon and inorganic carbon respectively. Total organiccarbon is the difference of total carbon and inorganic carbon.Volatile Fatty AcidsVolatile fatty acids, total carbohydrates and tamiin-lignin were tested to assess theavailability of substrate and its conversion in the reactors. Volatile fatty acids have provento be a good indicator of the condition of an anaerobic system. VFAs are among the mostimportant intermediate products formed during anaerobic digestion. They are thesubstrates of methane production and sulphur reduction and they influence the acid-baseequilibrium of the fermentation liquor. Changes in the individual WA concentrationsreflect changes in the microbial activity of the process.Volatile fatty acids (formic acid, acetic acid, propionic acid, n-butyric acid and ibutyric acid) and lactic acid were measured on a Varian 5000 Liquid Chromatograph. 100microlitres of filtered samples (C 18 Sep-Pak and Millex-HV 13 45 urn cartridges, WatersChromatography, Milford MA) passed through an Aminex Ion Exclusion HPX-87Horganic acid column (300 mm x 7.8 mm I.D.) supplied by Bio-Rad Laboratories(Richmond, CA). A Micro-Guard Cation H (Bio-Rad Laboratories) guard cartridgeprotected the column from contamination. The column was operated at 20 °C and waseluted with a 0.0 13 N sulphuric acid solution at 1.0 mi/mm after the method of Schneideret a!. (1987). Volatile fatty acids were assayed on a Waters Ion Chromatograph equippedwith a 486 Tunable Absorbance detector at 210 nm.85The liquid chromatograph was calibrated by means of an external standard whichwas prepared to profile the samples. The system demonstrated near linearity for the rangeof calibration standards which were used. Samples obtained from the reactor system werestored in glass vials and immediately frozen until convenient for analysis.Suiphide, Sulphate, Suiphite and ThiosuiphateSulphate, sulphite, thiosulphate and total dissolved suiphide concentrations weremonitored to assess the efficacy of the sulphur management technique. Total dissolvedsuiphide and sulphate analyses were performed on centrifuged (35000 rpm for 15 minutes)fresh feed, feed plus nutrient solution sampled after one day from the room temperaturefeed tank, each of the three acid reactors and the three methane phase reactors. Thesuiphide and sulphate analyses employed a Hach 2000 spectrophotometer following themethylene blue and barium sulphate turbidimetric methods respectively. The methyleneblue suiphide test is based on the ability of hydrogen suiphide to convert N,N-dimethyl-pphenylenediamine directly to methylene blue in the presence of potassium dichromate, anoxidizing agent. The intensity of the blue colour measured at 665 nm is proportional to thesuiphide concentration. Sulphate was measured at 450 nm via its quantitative precipitationwith barium chloride.The very labile nature of suiphide required that these tests be performedimmediately, without sample storage. Samples were assayed daily until steady state wasapparent. From the standpoint of stable sulphide and sulphate levels, steady state wasgenerally achieved in less than two weeks from the start of an experiment.Suiphite, thiosuiphate, suiphide and sulphate were assayed on a Waters IonChromatograph equipped with a 486 Tunable Absorbance detector and a Conductivitydetector using the Waters Sulphur Speciation method A-ill. A 4 mM sodium hydrogenphosphate solution constituted the mobile phase which was pumped at a flow rate of 1.0mi/mm at approximately 20 °C through an IC-Pak Anion 4.6 x 50 mm column.86Samples were centrifuged (35000 rpm for 15 minutes) and frozen in glass vialsuntil the analyses were performed on the ion chromatograph. 10 mM mannitol was addedto stabilize the diluted samples according to the Waters method. Samples were thenfiltered with a Cl 8 Sep-Pak cartridge (Waters Chromatography, Milford, MA) in order toremove UV absorbing phenolic substances. This step was followed by 0.45 urn filtration(Millipore Millex-HV1 3) to remove extremely small particulates.Gas AnalysesMethane and carbon dioxide were measured to quantifj the activity of the methaneproducing bacteria. The hydrogen sulphide gas concentration was assessed during the gasscrubbing and recycle experiments in order to measure the effectiveness of the hydrogensulphide scrubbers. Methane, hydrogen, and carbon dioxide were measured on a Vista6000 gas chromatograph using a Porapak N 50/80 12 ft by 1/8 inch (3.66 mx 3.2 mm) SScolumn with the thermocouple detector set at 210 °C. The column was held at 50 °Cduring the period of peak elution. Argon or helium was used as the carrier gas at 20mllmin. The calibration gas was of the following composition: 60.04% methane, 25.00%carbon dioxide, 9.97% hydrogen, 3.00% nitrogen and 1.99% oxygen. The methanecontent of the methane phase reactor head space was analyzed at least twice perexperiment.Hydrogen suiphide was measured for each of the BCTMP/TMP gas scrubbingexperiments on a Vista 6500 gas chromatograph using a Carbopak BH-T (15 inch by 1/8inch Teflon) column and flame photometric detector. Column temperature was 450 °C.Helium was used as the carrier gas at a flow rate of 30 mllmin. The hydrogen suiphidecontent of the methane phase reactor head space was analyzed at least twice per gasstripping experiment.875.8 Plan of ExperimentsThe experiments were performed to assess three methods of sulphur managementin the anaerobic treatment of the mechanical pulping effluent which was obtained fromQuesnel River Pulp Co. The impacts of acid phase reactor pH, molybdate addition, andsuiphide gas stripping were examined in these experiments. Each of these sulphurmanagement strategies was tested over a range of three HRTs at both mesophilic (35 °C)and thermophilic (55 °C) temperatures using the two phase anaerobic reactor systemswhich were described in section 5. The effects were evaluated in terms of sulphatereduction, total dissolved sulphide concentration, suiphite concentration, thiosulphateconcentration, TOC removal, gas production, the VFA profile, pH and sodium carbonateconsumption.A summary of the sulphur management experiments which were conducted at 35°C is given in Table 1. In the first sulphur management strategy assessment, the acid phasereactor (APR) pH was varied from 5.5 to 8.0. Following a 40 day shut down, system restart was then monitored at the APR pH 7.5 setting. Molybdate additions to the feed tank,at levels up to 1.0 mM, were then assessed. The 1.0 mM molybdate experimental run wasrepeated for three different Quesnel River Pulp effluent batches in an attempt todemonstrate the reproducibility of the results. Iron, as FeC136H2O, was then added at100 and 200 mg/i levels in an attempt to augment the 1.0 mM molybdate effect oflowering the dissolved sulphide concentrations. Gas scrubbing and stripping experimentswere then conducted to remove from the methane phase reactor the suiphide which wasformed. Two different effluent batches were fed to the reactor system for theseexperiments. This was done to examine the reproducibility of the runs and to consider theeffects of high and low influent sulphate concentrations.The 55 °C thermophilic sulphur management experiments which were performedafter the mesophilic runs are summarized in Table 2. Fewer experiments were conducted88at this temperature compared to the mesophilic work since the mesophilic results gavesome insight into the operating levels of the sulphur management variables.Table 1: Sulphur Management Experiments Performed at 35 °CExperiment APR Mo Scrubbing EffluentNumber pH [mMl of MPR Batch Comments1 5.5 0 none 1, 2 uncontrolled pH2 6.5 0 none 33 7.0 0 none 44 8.0 0 none 45 7.5 0 none 46 7.5 0 none 5 start-up after 40 daysdown7 7.5 1.0 none 6 effluent batch 18 7.5 0 none 6 0 molybdate control9 7.5 0.1 none 610 7.5 0.5 none 611 7.5 0.75 none 612 7.5 1.0 none 7 effluent batch 213 7.5 1.0 none 8 effluent batch 314 7.5 1.0 none 8 +100mg/i____________________FeCl*6H,O15 7.5 1.0 none 8 +200mg/iFeCl *6H,O16 7.5 0 water scrubber 9 gas scrubbing control17 7.0 0 FeC13 scrubber 9 effluent batch 118 7.0 0 FeCl3 scmbber 10 effluent batch 2(low sulphate)19 7.0 0 FeC13scrubber 10 effluent batch 2+ NaSO489Following an approximate 1 C°/day 35 °C to 55 °C mesophilic to thermophilictemperature adjustment, one month of continuous operation was allowed to pass toacclimate the bacteria at 55 °C before any sampling was resumed. Then, the acid phasereactor pH was adjusted from 7.0 to 8.0, molybdate was added at 0.5 and 1.0 mM, andgas scrubbing and stripping experiments were performed, all at 55 °C.Table 2: Sulphur Management Experiments Performed at 55 °CExperiment APR Mo Scrubbing EffluentNumber pH FmMI of MPR Batch Comments20 7.5 0 none 11,12 35°C-55°C@1C°/d21 7.0 0 none 1322 8.0 0 none 1323 7.5 0 none 1324 7.5 1.0 none 1425 7.5 0 none 14 0 mM Mo control26 7.5 0.5 none 1427 7.0 0 water scrubber 15 gas recycle control28 7.0 0 FeC!3 scrubber 15 H2S gas scrubbing90Chapter 6Characterization of the BCTMPITMP EffluentThe effluent obtained from the BCTMP/TMP operation at Quesnel River Pulp wasan opaque, orange-brown liquid. of this effluent, although highlyvariable, was within the range reported in the literature for most parameters. Table 3summarizes the characteristics of this effluent. Most notable from Table 3 were the highsulphate, COD, TOC and acetate concentrations, suggesting that while anaerobicprocessing may be a viable treatment option, the sulphate levels of up to 1565 mg/l may beproblematic. The COD:S04 ratio of the Quesnel River Pulp effluent was approximately5:1 or roughly 2.5:1 biodegradable COD:S04. This ratio is dramatically less than theminimum ratio of 10:1 to 100:1 for uninhibited anaerobic treatment, which was specifiedby a number of investigators.One notable departure from the literature is the high concentration of VFAs whichwere measured in the Quesnel River Pulp effluent. It is not known whether or not theconcentrations of these acids were artifacts of the time delay resulting from thetransporting the effluent or due to some unusual aspect of the Quesnel River Pulp Co.pulping process. The concentrations of the VFAs in the effluent were stable howeverduring the 3 °C storage for the duration of these experiments. In contrast with BeakConsultants (1986) who claimed that such wastewaters are poorly hydrolyzed, themaximum total VFA concentration from Quesnel River Pulp was approximately triple thatreported elsewhere. This fact was reflected in the COD:BOD5ratio of the effluents, wherethe Quesnel River Pulp effluent ratio was approximately 2.0, compared to almost 3.0reported elsewhere (Andersson et aL, 1985). The lower ratio indicates that this effluentwas more biodegradable than similar wastewaters generated by other CTMP mills.91In spite of sodium suiphite being used as a reagent in the pulping process, onlysmall suiphite concentrations were detected in the effluent. These sulphite concentrationsin the feedstock were only a very small percentage of the influent sulphate levels. Thissuggests the usefulness of considering sulphate to be the principal parameter with which tomonitor the efficacy of the SRB. -No hydrogen peroxide was ever detected in this effluent in spite of its additionwith DTPA by the mill. This deserves an explanation in light of the excellent work byWelander (1989) in particular regarding the management of hydrogen peroxide foranaerobic treatment of similar CTMP effluents. Perhaps the zero peroxide levels were dueto its consumption by the inorganic sulphur species during transport from the mill to thelaboratory. The low sulphite, thiosuiphate and sulphide concentrations may point to theiroxidation by peroxide to sulphate.The very small values of the acute fish (rainbow trout) toxicity bioassay(approximately 1% by volume) and the Microtox EC5O assay of less than 4%, indicatethat this effluent would exert a significant negative impact if released untreated to thereceiving waters. These toxicity parameters as well as the DTPA concentration, and the50% IC (MPB) (the concentration which inhibits 50% of the methanogenic activity) ofless than 10% for similar effluents (Sierra-Alvarez et aL, 1993), indicates that the QuesnelRiver Pulp effluent would be difficult to treat by biological means in general and byanaerobic means in particular.92Table 3: Characterization of the BCTMP/TMP EffluentBCTMP/TMP CTMP Effluent TMP EffluentParameter Effluent Reported in Reported in1mg/Il in this study the literature the literatureCOD 2520-7930 2100-13000 2650-9000BOD5 15003500* 1000-4500 1000-4700TOC 1065-3560 650 980-2700TSS 1400-3570 180-5000 40-1360Total VFAs 430 - 2720 400 - 800 200 - 400Formate 0 - 340 - -Acetate 350-1465 1500 0-235n-Butyrate 20 - 155 - 0 - 20Propionate 20 - 320 - 0 - 20Lactate 40 - 440 - -pH 5.5 - 8.0 5.0 - 10.0 4.5 - 8.4SO42- 525 - 1565 200 - 1590 200 - 700SO32- 10 - 30 0 - 225 200S203 0 - 10 - -S2 0.7-3.3 - -WoodExtractives 36** 25-550 2-20H20 0 50 - 1000 0 - 100DTPA <300*** 20-500 -96 h LC5O [vol %] 0.9**** 0.3- 8.0 0.5 - 2.4EC5O[vol%] 0.6-3.8 - -50%IC(MPB)[vol%] - 1.5-8.2 11.5-13.7Temperature°C 5070*** 50-80 40-60* Quesnel River Pulp effluent assayed by Dubeski, 1993.** Quesnel River Pulp effluent assayed by McCarthy et aL, 1990.Quesnel River Pulp effluent assayed by Puar, 1991 (personal communication).Quesnel River Pulp effluent assayed by MacLean et aL, 1990.93Chapter 7Sulphur Management Strategy 1:Elevated Reactor pH7.1 OverviewSection 2 explains the effects of pH on the ionic form and toxicity of suiphide tomethane producing bacteria.Section 3 describes the results of the acid phase reactor pH experiments whichwere conducted at 35 °C.Section 4 describes the results of the acid phase reactor pH experiments whichwere conducted at 55 °C.Section 5 summarizes the conclusions from these experiments.7.2 BackgroundThe reactor pH affects sulphate reduction in two principal ways:i) by influencing the metabolic rates of the SRBs, andii) by governing the relative amounts of the suiphide species (H2S, HS- or s2-) presentwhich are a function of pH.The toxicity of sulphide to the MPB depends mainly upon the concentration of freehydrogen suiphide in solution and not on the level of total suiphide (Schiegel, 1969, Kroissand Plahl-Wabnegg, 1983, Isa et aL, 1986, Sarner et al., 1988, McCartney andOleszkiewicz, 1990). According to Neal et a!. (1965), the free hydrogen sulphide effectcould be due to the greater ability of an uncharged molecule to pass through a cellmembrane by ordinary diffusion compared to an ion.The chemical equilibrium of the various sulphide species is a function of pH.Depending on the pH of the medium, the suiphide will be present as hydrogen sulphide(H2S), bisulphide ion (HS-) or sulphide ion (S2j. Since the inhibition of methanogens is94due to un-ionized hydrogen suiphide the concentration of which is proportional to thereciprocal of the exponent of pH (Hilton and Oleszkiewicz, 1988), consequently addingalkali to increase the pH and shift unionized sulphide to the less toxic HS- has beenproposed by a number of investigators (Hilton and Oleszkiewicz, 1987, Nanninga andGottschal, 1988, Sarner etaL, 1988,Lee et a!., 1989).The portion of total dissolved suiphide which is present as free H2S is a functionof pH and is calculated using equation 1 (on page 67). The percentage of un-ionizedhydrogen sulphide is 90% at pH 6.0 and drops sharply at pHs greater than 6.4 to roughlyequimolar concentrations ofH2S and HS- with negligible S2 at neutral pH. At pH 7.5,approximately 20% of the total sulphide present in solution exists as un-ionized sulphide.At an APR pH of 8.0, this fraction drops to about 10%.According to Hilton and Oleszkiewicz (1988), the SRB are more sensitive to thetotal dissolved sulphide concentrations than are the MPB. Since raising the pH decreasesthe free hydrogen sulphide concentration but leaves the total dissolved sulphide constant,the MPB should demonstrate decreased suiphide inhibition at elevated pH levels and bebetter able to compete for substrate with the SRB. The practical upper pH limit dependson the pH effect on the metabolic activities of the bacterial groups. Although the work ofOleszkiewicz and co-workers challenges this, the MPB are almost universally cited ashaving a narrow, near neutral pH range. The SRB favour more alkaline conditions.Some authors (Sarner et at., 1988, Lee et aL, 1989) have advocated adjusting thedigester pH to approximately 7.5 in order to shift the sulphide species away from theundissociated form. Others (Hilton and Oleszkiewicz, 1987) quoted a higher pH: a rangeof 8.0 to 8.4 where biological methane formation can proceed uninhibited in the presenceof high total suiphide but low un-ionized sulphide concentrations. Increasing the pH levelsignificantly above neutral can lead to inhibition of the MPB directly and such inhibitionplaces an upper limit on pH at 8.0, while McCartney and Oleszkiewicz (1990) have foundno MPB inhibition at this level.95While the pH affects the treatment system, the treatment system also affects thepH. Middleton and Lawrence (1977), Laanbroek and Pfenning (1981), Maree andStrydom, 1985 and Oleszkiewicz and Hilton (1986) have reported that sulphate reductionstabilizes pH at a relatively high level and that higher sulphate removal efficiencies led tohigher alkalinity values: This is-due-to-both the -consumption of organic acids by the SRBand to production of hydroxyl ions as sulphate is reduced.pH regulation of the acid phase or pre-reactor is simple and could improve theextent of sulphate reduction and TOC removal by the SRB while minimizing the inhibitoryeffects of the sulphide by product. Such enhanced sulphate reduction should result inimproved carbon removal by the SRB. There are drawbacks to this simple approachhowever.a) The cost of chemicals to maintain an elevated reactor pH is high (Nanninga andGottschal, 1987, Jensen et al., 1988, Sarner et at., 1988). After treatment, the alkalineeffluent must also be neutralized prior to other downstream biological treatment or to finaldischarge to the receiving waters, further increasing chemical consumption. Decreasingthe pH to the neutral range would also have the effect of releasing hydrogen suiphide gasfrom solution since the portion of free hydrogen sulphide increases with decreasing pHand H2S is distributed between the aqueous and vapour phases.b) High sulphide levels will remain in the effluent which will raise the COD and also createliquid handling difficulties downstream.c) Ionized sulphide is not partitioned between the liquid and gaseous phases. Hydrogensulphide is alone among the sulphide forms which partitions between the liquid and thegaseous phases. The total dissolved sulphur concentration will be decreased at the acidicpH levels where hydrogen suiphide predominates. Therefore the gas stripping whichoccurs in anaerobic treatment systems as a natural consequence of methane and carbondioxide production will not help to decrease the sulphide levels in the alkaline digesterliquor (Nanninga and Gottschal, 1987).967.3 Results: Effect of Acid Phase Reactor pH at 35 °Ca) SulphateAs the acid phase reactor pH was increased from 5.5 to 8.0, the sulphate reductionefficiency ([1-final S04/initial SO4]* 100) appeared to be unaffected (Table 4). Regressionanalysis also confirms the insignificant effect of the APR pH on sulphate reduction withthe significant variable being HRT. (Regression analysis was performed for each of theexperiments. The effects of the two manipulated variables, the sulphur managementstrategy and the HRT, were assessed using stepwise regression.) The exceptions to thistrend are the lower sulphate removal efficiencies which were observed at the smallest HRTwith increased APR pH. (It should be noted that the sulphate reduction efficiency for themethane phase reactors is an overall efficiency and it includes the sulphate removed in theacid phase reactors.) This is in agreement with Hilton and Oleszkiewicz (1988) who foundsulphate reduction to be inhibited in proportion to the total dissolved suiphideconcentration in contrast with literature reports of enhanced SRB activity at increased pH(Minami et al., 1988). Minami et at. (1988) found approximately 40% suppression ofsulphate reduction at acidic pHs between 6.2 and 6.8, whereas a neutral to slightly alkalineliquor supported almost complete reduction to suiphide.The sulphate reduction noted in this work was considerable in the acid phasereactors, ranging from 39 to 84%. Additional sulphate reduction occurred in the methaneproducing reactors but the effluent was never reduced down to zero sulphate levels. Aresidual sulphate concentration typically ranging from 150 to 250 mgIl remained in the exitstream of the methane phase reactors. The residual sulphate may have been be due to:1) sample handling (pH adjustment, centrifugation and dilution) leading to the oxidation ofsome of the dissolved suiphide back to sulphate,2) mixing in the reactors, leading to some short circuiting of the incoming sulphateconcentrations, or973) nutrient limitations of the SRB. Puhakka et al. (1990) for example observed the SRB tobe limited by a deficiency of complex (greater than 2 carbon) VFAs.There appears to be no HRT effect at the low influent sulphate concentrations.With increased pH and increased influent concentrations, the smallest HRT reactordemonstrated a much lower capacity to reduce sulphate than did the intermediate and longHRT reactors. Unfortunately, since the acid phase reactor pH and the influent sulphateconcentrations are confounded, no independent determination of the effects of pH andsulphate concentration can be made.Table 4: Effect of Acid Phase Reactor pH on Sulphate Reduction at 35 °CMean Sulphate Reduction Efficiency L%] ± CVFeed Acid/Methane Phase Reactor HRT jdays]Acid Phase S042±SD APR MPR APR MPR APR MPRReactor pH [mg/I] 0.2 0.4 0.4 0.8 0.6 1.255 600±85 64±9 73±8 68± 14 87± 12 71± 12 90± 176.5 642±85 74± 12 75±9 67±9 75±6 75± 13 78±77.0 1069±55 40±7 49± 12 82±5 89±6 77±25 87± 1075 1216±64 47±9 79±9 78± 10 86± 16 83± 13 88±218.0 869±43 39±8 46±10 78±15 84±7 84±5 85±5b) Total Dissolved SuiphideSince sulphide is the product of the reduction of the inorganic oxidized sulphurcompounds, principally sulphate in this effluent, the trends for sulphate removal should bereflected in suiphide levels. This was as expected and the effect of acid phase reactor pHon total dissolved suiphide concentrations is summarized in Table 5.98The sum of dissolved hydrogen suiphide, bisulphide and suiphide levels increasedwith increasing acid phase reactor pH and with influent sulphate concentration. Thesulphide concentration was smallest at the lowest HRT setting but there appeared to be nosignificant difference between the intermediate and the longest HRT.Total dissolved concentrations were very low- in the- feed (2 to 3 mg/i) andincreased to levels exceeding 100 mg/i for this effluent batch. Later control experimentsdemonstrated the sulphate and sulphide concentrations to vary significantly from batch tobatch of effluent. This variability is shown in Table 66 which summarizes the performanceof the reactor system at an APR pH of 7.5 for three different effluent batches.The experiments were conducted in the following order (as acid phase reactorpH): 5.5, 6.5, 7.0, 8.0 and 7.5. Consequently, the acid phase reactor pH was confoundedwith the time of the experiment. This was considered to be unavoidable since it wasconsidered to be necessary in order to minimize the abrupt changes to the wastewatertreatment microorganisms. However, the acid phase reactor pH of 7.5 experiment was runout of sequence. Since those results fit the general trends, the importance of theexperimental sequence appears to have been small.Expressed as a percentage of sulphate sulphur which was reduced in the reactors,the recovery of total dissolved sulphide sulphur increased with increased pH, from only10% at the uncontrolled pH setting to a maximum of 50% sulphur recovery at the acidphase reactor pH of 8.0. Consequently, the sulphate concentration would appear to be themore reliable parameter.The suiphide levels increased with increased influent sulphate concentrations andwith increased APR pH. The total dissolved suiphide could be predicted from the acidphase reactor HRT (IIRTAPR) and APR pH according to:S2pR = -190 + 3J*pHp+ 57*HRTAPRr=O.83 p=O.OOl99S2MpR = -175 + 31 *pHp + 16*HRTMpRr=O.86 p=O.000The units for the parameters which appear in these and all of the following regressionequations are those used throughout the experiments: suiphide in mg/i, TOC and sulphateremovals in %, HRT in days. These regression equations indicate that, at p considerablysmaller than 0.05, there is good fit to the data and with r, the multiple Pearson correlationcoefficient, close to 1.0, the relationship approximates linearity. With positive coefficientsfor the APR pH and the HRTs, increasing either of these parameters results in an increasein the total dissolved suiphide level.Part of the effect of APR pH may be due to the declining fraction of hydrogensuiphide with increasing pH. Hydrogen sulphide is unique among the suiphide speciesbecause it is distributed between the liquid and gaseous phases as described by Henry’slaw. On the other hand, the ionic suiphide species of bisuiphide and sulphide remain in theliquid phase. Thus lowering the pH provides a means of escape of dissolved suiphide fromsolution as hydrogen suiphide gas. Raising the pH to shift the sulphide species to theionized and less toxic form results in a retention of the total dissolved sulphide in solution.For this approach of sulphur management to be useful, the decrease of theinhibitory effects due to the smaller fraction of hydrogen sulphide present in solution mustbe greater than the product of the increased concentrations of ionized sulphide species andtheir respective toxicities.c) Suiphite and ThiosuiphateAll of the samples were spoiled before sulphite and thiosuiphate testing could beperformed. Consequently, no other oxidized inorganic sulphur species can be reported onnor could a sulphur balance be performed for this set of experiments.100Table 5: Effect of Acid Phase Reactor pH on Total Dissolved SuiphideConcentration at 35 °CMean Total Dissolved Suiphide Concentration ± SD [mg/i]Feed Acid/Methane Phase Reactor HRT [days]Acid Phase s2± SD APR MPR APR MPR APR MPRReactor pH [mg/i] 0.2 0.4 0.4 0.8 0.6 1.25.5 2±0.1 14±1 16±2 16±1 16±3 11±3 15±16.5 3±0.1 10±1 24±1 13±1 25±3 21±2 19±17.0 3±0.2 25±6 40±3 82± 10 87± 12 67± 11 77± 157.5 3±0.2 40±5 59±7 74±6 79±4 76±7 73±58.0 3±0.1 46±9 59± 14 112±8 105±9 102± 11 101± 10d) Wastewater Treatment EfficiencyA summary of the removal of total organic carbon is presented in Table 6. Itshould be noted that the TOC removal efficiency for the methane phase reactors is anoverall efficiency and it includes the TOC removed in the acid phase reactors. The TOCremoval efficiencies were very low. These experiments demonstrated that a considerablefraction of the dissolved organic compounds of this effluent was not anaerobically biodegradable under the conditions examined. Such low removal rates of organic compoundsare in general agreement with the literature. Beak Consultants (1986) for example, quotedthe COD:BOD5ratio of CTMP effluent as typically 2.8:1. Thus, complete BOD5 removalwould correspond to a COD treatment efficiency of only 36%.The TOC treatment efficiency ([1.0 - final TOC/initial TOCj* 100) increased to amaximum of approximately 60% at an acid phase reactor pH of 7.5. Considering thegenerally high recalcitrance of this effluent, such a high treatment efficiency is indicative ofa wastewater batch which was much more degradable than most. The location of this101optimum was likely due to a declining performance of the MPB above this pH and adeclining performance of the SRB below this pH.The effect of APR pH on the TOC removal efficiency was found to be as follows:%TOCrApR -50 +8*pHp + 27*HRTApRr=0.76 p=O.OOS%TOCrMpR 75 + 12 *pHp +23 *HRTMpRr=0.82 p=O.OOJThese regression equations indicate that increasing the APR pH and/or the HRT, resultedin increased wastewater treatment efficiencies.Lo et al. (1991) also found pH to exert a significant impact on the removal oforganic compounds. They observed a neutral pH to be the treatment environment whichproduced the greatest removal efficiencies when anaerobically treating a BCTMP effluent.Any decrease in sulphide inhibition of the MPB at pH levels greater than 7.5 may be offsetby the lower metabolic rates of the MPB outside of their optimal pH range. While theperformance summarized here demonstrated no significant time trends over the course ofsampling, it can be speculated that longer term experiments at these alkaline conditionsmight have allowed the favoured SRB population to increase and consequently to improvethe TOC removal efficiency.Not controlling the acid phase reactor pH resulted in a mean pH of 5.5, far belowthe optimal ranges of both the MPB and the SRB. Both this setting and a mean acid phasereactor pH of 6.5 proved ineffective with respect to TOC removal.TOC removal increased with increasing HRT from target settings of 0.6 to 1.8days. There appears to be a benefit in increasing the HRT from 0.6 to 1.2 days butsignificant gains in TOC removal were not realized at the 1.8 day HRT compared to the1.2 day HRT setting.The acid phase reactors contributed roughly half of the total TOC removals inspite of being only one third of the total treatment volume. If the TOC removal efficiency102was simply HRT related and independent of the substrate composition (zero order kineticsas would be characteristic of high concentrations of slow to degrade organic constituents),only a third of the total TOC removal in the acid phase reactors would be anticipated.Thus the acid phase reactors were more efficient in terms of TOC removal per unit ofreactor volume than the methane phase reactors. This may have been due to a number offactors.1) The more readily degradable substrate was removed first in the acid forming reactors,leaving the more recalcitrant substrate to be removed downstream in the methane formingreactors.2) Inhibition of the waste treatment microorganisms was more pronounced in the methanephase reactors where the sulphide concentrations were higher.3) The UASB upflow recycle configuration of the acid phase reactors may have been moreeffective than the UASB/flxed film hybrid reactors in the methane forming stage.Table 6: Effect of Acid Phase Reactor pH on TOC Removal at 35 °CMean TOC Removal Efficiency [%] ±5% CVFeed Acid/Methane Phase Reactor HRT [days]Acid Phase TOC APR MPR APR MPR APR MiRReactor pH [mg/I] 0.2 0.4 0.4 0.8 0.6 1.25.5 1065 4 5 9 16 8 136.5 1050 5 14 7 16 16 317.0 2415 11 14 7 26 14 237.5 3235 20 24 39 55 36 638.0 3175 11 23 24 41 26 39103Due to the large batch to batch variability of this effluent, TOC removals would bedifficult to predict simply from selecting reactor conditions of these experiments.Therefore, while comparing the treatment effects within a wastewater batch can be madewith some confidence, the feedstock variability confounds the ability to draw more generalconclusions. -e) Gas ProductionThe mean daily gas production rates followed the same trends as for the TOCremovals. The mean gas production rates are summarized in Table 7. Consistent with theTOC results, the gas production rates increased with increased HRT and exhibited amaximum at the acid phase reactor pH 7.5 setting. APR pH and HRT influenced gasproduction according to:gas (i/day) = -0.9 + 0.2 *pH1j + 3.5 *HRT,jpr=0.72 p=O.OL3Both APR pH and HRT promoted gas production but the relationship demonstrated somedeviation from linearity.While the TOC and gas production rate trends were in general agreement, theactual carbon removals were not reflected in the methane gas production rates which werehighly variable and much less than those anticipated from theory.Low methane production rates are not uncommon when treating effluents from thepulp and paper industry. These effluents are at times complex, inhibitory or recalcitrant.Pichon (1987), Welander (1987) and McFarland and Jewell (1990) similarly observed themethane rates to be diminished under conditions of high sulphur loading. From their lowermethane yields in the presence of high acetate concentrations, they concluded thatcompetitive inhibition was not the only mechanism causing methanogenic inhibition. BeakConsultants (1986) found the methane yields to vary between 0.1 and 0.3 m3 CH4/kgCODr, averaging 0.16 m3 CH4/kg CODr, compared to the theoretical value of 0.35 m3104CH4/kg CODr. They conjectured that this difference could be ascribed to the SRBinhibiting the MPB via suiphide accumulations. Since 2.0 g COD are required to reduce1.0 g of sulphate or suiphite sulphur, the methane yield is decreased by 0.7 m3 CH4/kg Sreduced.The methane content of the gas was observed to vary between 38 and 58%. Thebiogas composition demonstrated considerable variability between sampling days. Ingeneral, the methane content increased with increased HRT and with increased acid phasereactor pH to a maximum of 58% at pH 7.5. The pilot scale anaerobic treatment study ofQuesnel River Pulp effluent conducted by Beak Consultants (1986) also found themethane content in the biogas to vary a great deal, from 40 to 80% methane, median 65%.The higher methane levels obtained in the Beak Study suggest that the methanogens wereless inhibited than in this work. This may have been due to an inhibition of the MPB by theundiluted effluent in this study. The Beak results were obtained with a 1:1 dilution of theQuesnel River Pulp effluent with cooling water.Table 7: Effect of Acid Phase Reactor pH on Gas Production Rate at 35 °CMean Gas Production Rate ± SD [lid]Acid Phase Acid+Methane Phase Reactor HRT [days]Reactor pH 0.6 1.2 1.85.5 1.2±0.2 3.4±0.8 2.6±0.46.5 0.9±0.4 4.5±0,8 4.2±0.97.0 1.4 ± 0.3 3.5 ± 0.3 3.3 ± 0.47.5 1.6 ± 0.4 3.5 ± 0.8 7.3 ± 1.28.0 1.1 ± 0.1 2.2 ± 0.1 3.1 ± 0.61051) Volatile Fatty Acids:Graphs depicting the individual volatile fatty acid concentrations as functions ofthe acid phase reactor pH are presented in Figure 3. A brief discussion of those trendsfollows.Formic Acid: Formic acid is a substrate directly metabolizable by the MPB andinfrequently by the SRB (Phelps et al., 1985, Battersby, 1988). In these experiments, thefeedstock formic acid concentrations were small and were decreased, likely by the IvIPB,to near zero levels at the larger HRTs in the methane phase reactors. The simple linearregression equation describing the effects of APR pH and HRT on formate levelsdemonstrated a poor linear fit (r = 0. 64 p = 0.090), indicating that these two variables arenot sufficient to adequately predict trends.Acetic Acid: According to the literature, acetate is generated both by the sulphur-utilizingand also the non-sulphur-utilizing bacteria, For this anaerobic treatment system which wasoperated at moderate organic loading rates (4 to 18 kg COD/rn3day), the elevated acetateconcentrations which were measured were indicative of low methanogenic activity. Thismay have been caused either by inhibition or by the absence of methanogens or SRBwhich were capable of using acetate.The results obtained by Hilton and Oleszkiewicz (1986), Rinzema and Lettinga(1988), Widdel (1988), Parkin et a!. (1990) and Maillacheruvu et aL (1993) demonstratedsimilar trends. Most of the unused organic substrate in these sulphide stressed treatmentsystems was present as acetate. Acetate utilization by the MPB was significantly affectedby the presence of sulphur compounds and sulphate reduction occurred by pathways otherthan acetate utilization. The predominance of acetate over the other VFAs in the MPRsand the apparent inability of the SRB to utilize acetate may help to explain the poorsulphate reduction rates in these reactors.106100875 E 1500 /Acid Phase Reactor pH Acid Phase Reactor pH500• 1000•400 . 800•< 300 600•7os:I:::____Acid Phase Reactor pH Acid Phase Reactor pH250-200- —•—Feedo 0.2dHRTAPR1‘d 150 ----.-- O.4dHRTAPR2100 -----0—0.6 d HRT APR 3500.4 d HRT MPR 1Acid Phase Reactor pHFigure 3: Effect of Acid Phase Reactor pH at 35 °C on Volatile Fatty Acids107Acetate was utilized early in this experimental program, indicating the initialpresence of acetoclastic methanogens or SRB in the microbial consortia. Since the acetatelevels were so high throughout the experiments, even the moderate levels of suiphidewhich were produced may have inhibited acetate uptake by either the MPB or the SRB. Inaddition, compounds other than sulphide such as resin acids or. DTP.A .may have beenresponsible for this chronic inhibition.At an acid phase reactor pH of 6.5, acetate was produced in the acid phase reactorand consumed in the methane phase reactor. The net acetate production in the acid phasereactor was independent of HRT, likely due to acetate production by the acetogens andconsumption by the MPB which were not washed out from the acid phase reactor. At theelevated acid phase reactor pHs, very high acetic acid concentrations were measured in theacid phase reactor.The acetate concentration increased with increased HRT in these reactors. Anincrease in the APR pH also led to increased acetate concentrations. The simple linearregression model to describe these effects is as follows:AcetateJR -3180 + 600*pHApR + 470 *HRTAPRr=0.75 pO.O24AcetateMpR = -2950 + 590*pfJApR 425*HRTMpRr=0.76 p=O.O2.2The positive HRTAPR coefficient indicates acetate generation in the APR and thenegative HRTp coefficient describes acetate uptake in the MPR.These high acetate levels indicate little significant inhibition of the SRB and nonSRB acetogens. The acetoclastic methanogens and SRB were quite ineffective. Even atthe longer HRTs, the net acetate consumption in the methane phase reactors was small,resulting in large residual acetate levels. The 0.4 day HRT methane phase reactordemonstrated no net consumption of acetate but rather continued to produce acetate. Theacid phase reactor pH of 7.5 resulted in the highest consumption of acetate in the methane108phase reactors. This optimum is consistent with the highest gas production rates and TOCremovals. It would appear that an APR pH of 7.5 favoured metabolic activity as measuredby a number of parameters.Propionic Acid: Propionic acid is also a substrate which is normally consumed by theSRB as well as by the non-sulphur utilizing acid producing bacteria to produce acetate andhydrogen. Slow rates of propionate degradation may be due to the accumulation of theproducts of degradation. Such slow rates are indicative of overloading or of inhibition.Given the low loading rates applied to the reactors over these experiments, overloadingcan be ruled out.The propionic acid concentrations which were measured in the fresh feed showedhigh batch to batch variability. Propionate was generated in the acid phase reactors andconsumed in the methane phase reactors but never down to zero levels. In general,propionate levels decreased with increasing HRT in the methane phase reactors but thehighly variable feedstock confounds any ability to identif’ trends of propionate withchanging acid phase reactor pH. Neither the APR pH nor HRT could account for thepropionic acid concentrations in the reactors (APR r2 0.072, MPR r2 0.244).Butvric Acid: High levels of n-butyric acid were generated at the acid phase reactor pHof 6.5 in both the acid phase and methane phase reactors. As the acid phase reactor pHincreased, these concentrations decreased to near zero levels at the largest HRT. Thesetrends are indicated by the negative coefficients for the regression equations:Butyratep = 1510 l8O*pHp 410*JJRTAPRr=O.78 p=O.O16Bu1yratep = 309238O*pHJ, 23O*TMPRr=0.74 p=O.O30Lactic Acid: Lactic acid is the preferred substrate of the SRB. Its availability helps toensure that sulphate reduction is not limited by organic substrate. Lactate was consumedto near zero levels at the acid phase reactor pH of 6.5. At the higher pH settings, there109was a net lactate consumption in both sets of reactors, leaving small residualconcentrations in the exit stream of the methane phase reactors. Neither the APR pH northe HRT could account for these trends in lactate concentration (APR r2 = 0.24, MPR r20.50). These detectable concentrations of lactate and propionate and butyrate in the exitstream of the MPRs -appear to refute any suggestion that incomplete sulphate reductionwas due to substrate limitation.g) Reactor pHThe acid phase reactor pH was varied at 35 °C from approximately 5.5(uncontrolled) to 6.5, 7.0, 7.5 and 8.0 by means of sodium carbonate addition. Theaverage pH values of the feed and of the acid phase and the methane phase reactors aresummarized in Table 8.The simple linear regression line describing the MPR pH in terms of the APR pHand the MPR HRT demonstrated an excellent fit:pHMpR = 2.9 + O.S*pHp + 0.5 *HRTMpRr=0.92 p=O.000Even though the caustic soda and sodium suiphite which were used in the CTMPcooking liquor produce an effluent which is neutral to slightly alkaline according to mostaccounts, the BCTMP/TMP effluent used for these experiments was slightly acidic. Theelevated VFA levels (Figure 3) suggest that this lower pH could have been due tohydrolysis of the dissolved carbohydrates by bacteria which already were in the milleffluent.Except for the experiments which involved the acid phase reactors without pHcontrol, the methane phase reactor pHs were all within a range tolerable by the MPB andthe SRB. van den Berg et at. (1976) and Patel (1984) observed the optimum rate of aceticacid conversion to occur between pH 6.5 and 7.5 with a sharply declining performanceoutside of this range. As the acid phase reactor pH decreased, the MPR declined as well.110The MPR pH was generally closer to neutrality than the pH of the corresponding APR.The smallest HRT always resulted in the lowest methane phase reactor pH, demonstratingthat acidogenesis continued in the MPR but larger HRTs enabled acid utilizing bacteria,such as the MPB, to consume substrate. Maree and Strydom (1985) noted that highsulphate removal efficiencies led to high alkalinityTable 8: Acid Phase Reactor and Methane Phase Reactor pH at 35 °CMean Reactor pH ± SDFeed Acid/Methane Phase Reactor HRT [days]Acid Phase pH ± APR MPR APR MPR APR MPRSDReactor pH 0.2 0.4 0.4 0.8 0.6 1.25.5 6.4±.2 5.9±.3 5.8±.i 5.3±.2 6.3±.3 5.4±.3 6.1±.16.5 6.7±.1 6,6±.2 6.0±.4 6.5±.1 6.7±.2 6.5±.2 6.6±.27.0 6.5±.1 7.0±.1 6.8±.2 7.0±.1 7.0±.2 7.0±.1 7.0±.17.5 6.4±.1 7,6±,1 6.9±.1 7.5±.1 7.3±.1 7.5±.1 7.4±.18.0 6.5±.1 8.0±.1 7.3±.3 8.1±.1 7.6±.1 8.0±.1 7.7±.1h) Sodium Carbonate ConsumptionThe 0.5 N sodium carbonate consumption was tallied for each experimental periodand divided by the total volume of effluent processed by the reactor system. These resultsare presented in Table 9. The consumption of sodium carbonate was small for the acidphase reactor in the acidic to neutral range. Sodium carbonate consumption increased withincreased acid phase reactor pH to a maximum of pH 7.5, not pH 8.0. This observationindicates that the net acid production in the acid phase reactor was a maximum at the pH7.5 setting.111Sodium carbonate consumption varied with the APR pH and the APR HRT asfollows:gNa2CO3/lfeed = -7.6 + ].6*HRTApR +l.2*pHpr=O.83 p=O.OOJThe APR pH was the variable which exerted the greatest effect on the carbonateconsumption.Table 9: Effect of Acid Phase Reactor pH on Carbonate Consumption at 35 °CMean Sodium Carbonate Consumption (gil feed]Acid Phase Acid Phase Reactor HRT (days]Reactor pH 0.2 0.4 0.65.5 0 0 06.5 0.12 0.16 0.267.0 0.45 0.45 0.527.5 1.81 3.59 2.848.0 1.49 2.90 3.00The carbonate consumption for the acid phase reactor settings of 7.5 and 8.0 wasfrom three to eight times that of the neutral setting. This indicates the high cost ofmaintaining even slightly alkaline reactor conditions.7.4 Results: Effect of Acid Phase Reactor pH at 55 °Ca) SulphateAs the acid phase reactor pH was increased from 7.0 to 8.0, the sulphate reductionappeared to decrease for all but the largest HRT. It should be noted that the sulphatereduction efficiency for the methane phase reactors is an overall efficiency and it includes112the sulphate removed in the acid phase reactors. The sulphate reduction was significant inboth the acid phase reactors and the methane phase reactors. Minami et al. (1988) found acomparable extent of sulphate reduction, markedly greater than 40% at 53 °C at analkaline pH and less than 40% sulphate reduction under even mildly acidic conditions.Overall, the sulphate reduction increased with increased HRT for these experiments.Neither the APR pH nor the HRT could account for the trend in sulphate reductionefficiencies (r2 = 0.31). Table 10 summarizes these sulphate trends.Table 10: Effect of Acid Phase Reactor pH on Sulphate Reduction at 55 °CMean Sulphate Reduction Efficiency j%J ± CVFeed Acid/Methane Phase Reactor HRT LdaysjAcid Phase S042±SD APR MPR APR MPR APR MPRReactor pH [mg/Il 0.2 0.4 0.4 0.8 0.6 1.27.0 1245±5 36±7 46±8 33±8 48±12 19±6 42±67.5 1411±2 32±7 40±6 36±7 43±9 34±5 50±48.0 1189±4 25±5 33±7 19±4 39±4 22±6 51±5Lower fractions of sulphate were reduced at 55 °C compared to the 35 °C work.Whereas at 35 °C, the sulphate reduction was considerable, ranging from 46 to 90%,under similar loading conditions thermophilic sulphate reduction reached a maximum ofonly 50%. Lepisto and Rintala (1993) also observed thermophilic sulphate reduction to beincomplete with effluent concentrations of 30 to 120 mg/l sulphate. McFarland and Jewell(1990) observed complete thermophilic sulphate reduction for a 10 day HRT. This lowerperformance by the thermophilic SRB could be due to a lower cell density arising from theat least partial elimination of mesophiles at thermophilic temperatures or from lower SRB113activity at elevated temperatures either inherently or due to a greater effect of inhibitingcompounds at elevated temperatures in the reactor liquor.b) Total Dissolved SulphideThere appears to be no relationship between total dissolved sulphide levels withthe acid phase reactor pH, quite unlike the 35 °C experiments. Similarly, there appears tobe no simple relationship between the sulphide concentration and the HRT. The squaredmultiple r values, that is the decimal fractions of the variation in the data which areaccounted for by the model, were only 0.23 for the APR and 0.16 for the MPR whenapplying simple linear regressions with APR pH and APR HRT as the independentvariables. Even at maximum sulphate reduction efficiencies of only approximately 50%(Table 10), the resulting mean total dissolved sulphide levels of the MPR samples stillexceeded 100 mgll in a majority of cases. A summary of this performance is presented inTable 11.Table 11: Effect of Acid Phase Reactor pH on Total Dissolved SuiphideConcentration at 55 °CMean Total Dissolved Sulphide Concentration ± SD [mg/i]Feed Acid/Methane Phase Reactor HRT [days]Acid Phase SO42-D APR MPR APR MPR APR MPRReactor pH [mg/i] 0.2 0.4 0.4 0.8 0.6 1.27.0 1245±5 91±7 113±8 80±9 112±6 58±11 116±77.5 1411±5 42±7 61±10 48±6 61±10 50±7 83±98.0 1189±4 59±4 85±5 62±4 109±5 62±4 13 1±51While the sulphate reduction efficiency decreased with increased pH, at 55 °C thefraction of total suiphide which was ionized and thus, not partitioned between the aqueous114and gaseous phases (where it could bubble out of solution) increased with increasing pH.These two factors appear to offset each other.c) Suiphite and ThiosuiphateSulphite concentrations were greater in the acid phase reactor and methane phasereactor samples than in the feed at the acid phase reactor pH 7.0 run. These results aresummarized in Table 12. The sulphite and thiosulphate levels were decreased withincreased APR pH. Increased APR HRT led to increased sulphite and thiosuiphateconcentrations in the APRs. The simple linear regression models to describe the trends ofsuiphite and thiosulphate in the APRs are as follows:SO3APR =152-19 *pHp + 7.5 *HRTpr= 0.74 p=O.O93S2O3ApR 148 -l9*pHp + 128 *HRTAPRr=0.82 p=O.O.36The simple linear regression equations were not adequate to describe suiphite andthiosulphate trends in the MPRs.In general, there was no significant difference between the acid phase reactor andthe methane phase reactor suiphite concentrations for these experiments. This suggeststhat suiphite functioned as an intermediate compound in the reduction of sulphate tosulphide. At the acid phase reactor pH levels of 7.5 and 8.0 however, the sulphite levels inany of the reactor samples were small and not significantly different from the feed. Eitherthere were no microorganisms capable of degrading this small residual sulphite at thosepH settings or the rate of sulphite formation was in equilibrium with its degradation sothat no net change occurred. HRT appeared to exert no effect on the suiphiteconcentrations for these experiments.115Table 12: Effect of Acid Phase Reactor pH on Suiphite Concentration at 55 °CMean Suiphite Concentration ± SD [mg/I]Feed Acid/Methane Phase Reactor HRT [days]Acid Phase SO32sD APR MPR APR MPR APR MPRReactor pH [mg/I] 0.2 0.4 0.4 0.8 0.6 1.27.0 11±2 26±2 21±3 25±4 19±3 33±5 33±57.5 9±2 4±2 4±1 5±1 6±1 5±1 5±18.0 9±2 9±2 9±2 9±2 15±1 10±1 11±2Thiosuiphate concentrations were also greater in the reactors than in the feedstockat all pH levels tested (see Table 13). Again, the acid phase reactor pH of 7.0 settingresulted in the highest concentrations of this intermediate. As the acid phase reactor pHwas increased, the mean thiosulphate concentration in the reactors generally declined. Ingeneral, thiosulphate concentrations appear to have increased with increased HRT.d) Wastewater Treatment EfficiencyIn general, the TOC removal efficiency was not influenced by either the pH or theHRT at the levels tested (APR r2 0.31, MPR r2 = 0.041). The mean TOC removalefficiencies were low (maximum removal at 24%) for these experiments, at odds with thetheories of enhanced biodegradability of the many recalcitrant organic compounds athigher temperatures. The small TOC removals (Table 14) were also reflected in the gasproduction rates (Table 15). It should be noted that the TOC removal efficiency for themethane phase reactors is an overall efficiency and it includes the TOC removed in theacid phase reactors.116Table 13: Effect of Acid Phase Reactor pH on Thiosuiphate Concentration at 55 °CMean Thiosulphate Concentration ± SD jmg/l]Feed Acid/Methane Phase Reactor HRT [days]Acid Phase S203 APR MPR APR MPR APR MPR±SDReactor pH [mgfl] 0.2 .0.4 0.4 0.8 0.6 1.27.0 9±3 62±11 42±19 37±11 51±23 73±18 118±227.5 5±3 12±3 47±3 54±5 52±11 72±6 82±38.0 5±1 9±4 19±5 45±11 92±7 65±16 84±23The low TOC removal efficiency, independent of HRT and reaching a maximum ofapproximately 24%, indicates that only a small fraction of the wastewater wasbiodegradable under these treatment conditions. There appears to be little additional TOCremoval in the methane phase reactors over that obtained in the acid phase reactors. Evenwith the HRT doubled in the methane phase reactors compared to the acid phase reactors,the methane phase reactor treatment microorganisms appear to be unable to significantlydegrade the organic constituents in the wastewater under these conditions. The exceptionsto these observations occurred at the acid phase reactor pH 8.0 experiments where theacid phase reactors functioned very poorly with respect to TOC removal but the methanephase reactors contributed to treatment efficiencies. This is probably due to the pH closerto neutral in the methane phase reactors.Similar to the mesophilic experiments, the TOC treatment efficiency increased to amaximum at an acid phase reactor pH of 7.5 (Table 14), likely due to a decliningperformance of the MPB above this value and a declining performance of the SRB belowthis value. Any decrease in sulphide inhibition of the MPB at elevated pH levels may beoffset by lower metabolic rates of the MPB outside of the optimal pH range. Longer term117experiments at these alkaline conditions may allow the SRB population to increase andconsequently to improve the TOC removal efficiency.Table 14: Effect of Acid Phase Reactor pH on TOC Removal at 55 °CMean TOC Removal Efficiency [%] ±5% CVFeed Acid/Methane Phase Reactor IIRT [days]Acid Phase TOC APR MPR APR MPR APR MPRReactor pH [mg/lI 0.2 0.4 0.4 0.8 0.6 1.27.0 3400 18 22 14 17 9 117.5 2930 13 20 16 18 19 248.0 3355 4 19 10 13 8 20e) Gas ProductionThe total gas production rates at 55 °C for these acid phase reactor pHexperiments were very small and demonstrated no relationship with either acid phasereactor pH or with HRT (r2 = 0.18). The average values for these experiments aresummarized in Table 15.Table 15: Effect of Acid Phase Reactor pH on Gas Production Rate at 55 °CMean Gas Production Rate 4- SD [lid]Acid Phase Acid+Methane Phase Reactor HRT [days]Reactor pH 0.6 1.2 1.87.0 0.10±04 0.24±.02 0.22±0.117.5 0.30±12 0.10±,04 0.12±058.0 0.09±,03 0.13±.05 0.16±,,031181) Volatile Fatty AcidsGraphs which depict the individual VFA concentrations for the acid phase reactorpH experiments at 55 °C are presented in Figure 4. A brief explanation of the VFA trendsfollows.Formic Acid: Even with fluctuating (150 to 340 mgll) formate levels in the feed, the acidphase reactors consistently reduced concentrations to near zero for all HRTs and pHsettings. Any residual formate was consumed in the methane phase reactors, presumablyby the MPB, to result in near zero formic acid concentrations. No adequate fit could bereached to describe formate using simple linear regression on the APR pH or HRT.Acetic Acid: Overall, acetate was generated in both the acid phase reactors and themethane phase reactors. The net acetic acid production was highest at the acid phasereactor pH of 7.5. Only poor fits were obtained to describe the effects of APR pH andHRT on the acetate concentration in the APR and MPR:AcetateApR = 4480 - 435*pHp + 500*HRTAPRr = 0.75 p = 0.085AcetateMpR = 529043O*pHp - 90*HRTMpRr=0.77 p=O.O7’OThese equations indicate that a higher pH decreases acetate levels and that a longer IVIPRHRT decreases the concentrations of acetate. Such high acetic acid concentrations areindicative of inhibited acetoclastic MPB and SRB.Propionic Acid: At the acid phase reactor pH 7.0 setting, propionate feed concentrationsexceeded those of the reactors, indicating a net consumption of propionate. In general,there was no significant difference between the propionate levels in the acid phase reactorsand the methane phase reactors. At an acid phase reactor pH of 7.5, this trend wasreversed with all of the reactor concentrations exceeding those of the feed, indicating a netproduction of propionate at this level. At the highest pH of 8.0, little difference in the feedand reactor concentrations could be established. It appears that neither the pH nor the119HRT affected the propionic acid levels in any simple way (APR r2 = 0.066, MPR r2 =0.099).Butvric Acid: At an acid phase reactor pH of 7.0, the butyrate concentrations in the acidphase reactor were considerably lower than in the feed. A net production of butyrate wasobserved in the methane phase reactors withbutyrate levels greater in the exit rather-thanin the intermediate streams. At a pH of 7.5, butyrate was produced in the acid phasereactors, resulting in concentrations greater than the feed. From this data, no generalstatement can be made concerning the effects of the acid phase reactor pH on butyrateconcentration in the methane phase reactor relative to those levels found in the acid phasereactors.Lactic Acid: Feed lactic acid concentrations were consistently reduced by the reactors.Little difference in lactic acid levels was evident in the acid phase reactors or the methanephase reactors at the levels of the acid phase reactor pH or HRT investigated. Lactic acidconcentrations were fhrther reduced in the methane phase reactors to near zero levels.g) Reactor pHFollowing the results at 35 °C, the acid phase reactor pH was set at 7.0, 7.5 and8.0 in order to assess the efficacy of retarding hydrogen suiphide inhibition by shifting thesuiphide species to the less toxic ionized form. The acid phase reactor pH was not testedat the uncontrolled or pH 6.5 levels since these settings were ineffective in the mesophilicinvestigation. Reactor pH was adjusted only in the acid phase reactor since the methanephase reactor had no direct provision for pH control. Table 16 summarizes the pHmeasurements recorded for these experiments and Table 17 documents the chemicalconsumption required in order to meet these pH targets.The methane phase reactor pH fell to near neutral with the largest pH drops fromthe acid phase reactor to the methane phase reactor at the largest acid reactor pH. This is1202000.400300 1:200 750o 500L 100 0250— :::::::::::::::.. 07.0 7.5 8.0 7.0 7.5 8.0Acid Phase Reactor pH Acid Phase Reactor pH500______________________________- 250400 200_Acid Phase Reactor pH Acid Phase Reactor pH500400 —— Feed0— 0.2dHRTAPR13000.4 d HRT APR 2200-----0----- 0.6 d HRT APR 3100-0- • 1.2dHRTMPR37.0 7.5 8.0Acid Phase Reactor pHFigure 4: Effect of Acid Phase Reactor pH at 55 °C on Volatile Fatty Acids121a result of the buffering capacity of the effluent and of the generation of VFAs whichaccumulated in the absence of very active MPB. The net VFA production was greatest atthe largest HRT, leading to the largest pH decrease to near neutral levels. There appearsto be no relationship between the methane phase reactor pH and the HRT for theseexperiments.Table 16: Acid Phase Reactor and Methane Phase Reactor pH at 55 °C:Mean Reactor pH ± SDFeed Acid/Methane Phase Reactor HRT [daysjAcid Phase pH ± APR MPR APR MPR APR MPRSDReactor pH 0.2 0.4 0.4 0.8 0.6 1.27.0 6.8±1 7.0±.1 7.0±.1 7.1±.1 6.6±.1 7.1±.1 7.1±.17.5 6.3±.1 7.5±.1 7.4±.1 7.5±.1 7.1±.1 7.6±.1 7.0±.18.0 6.2±.1 8.0±.1 7.3±.4 8.0±.1 7.6±.5 8.0±.1 7.4±.1h) Sodium Carbonate ConsumptionThe effect of acid phase reactor pH on carbonate consumption under thermophilicconditions is summarized in Table 17. As the pH was raised from 7.0 to 7.5, the sodiumcarbonate consumption increased, especially at the lower HRT. This performance can besummarized in a simple linear regression model as:gNa2CO3/lfeed = 4.] - O.46*pHp + 0.65 *HRTAPRr= 0.79 p=O.O50Increasing the acid phase reactor pH of 7.5 to 8.0 resulted in a significant declineof carbonate consumption. This non-linear behaviour could be attributed to:1) a decreased production of acids at the upper pH levels because of either lower activitiesof the acid producing bacteria or of the SRB,2) an increased consumption of VFAs with increasing pH by either the SRB or the MPB.122Table 17: Effect of Acid Phase Reactor pH on Carbonate Consumption at 55°CMean Sodium Carbonate Consumption [g/l feed]Acid Phase Acid Phase Reactor HRT [days]Reactor pH 0.2 0.4 0.67.0 0.88 0.89 1.337.5 1.56 1.46 1.768.0 0.54 0.78 0.407.5 Conclusions: Effect of Acid Phase Reactor pHAt 35 °C, changing the acid phase reactor pH from uncontrolled (resulting in a pHof 5.5) to 8.0 did not affect the degree of sulphate reduction, increased the total dissolvedsulphide concentration and improved the TOC removal efficiency. The sulphate reductionefficiency was considerable, up to 90% at the longest HRTs. This resulted in maximumtotal dissolved suiphide levels of approximately 100 mg/I, below the levels commonly citedby most investigators as inhibitory to the methanogens. The wastewater treatmentefficiency was markedly increased at an acid phase reactor pH of 7.5 with a maximumTOC removal of approximately 60%. However, elevated acetic acid concentrations andvery low rates of methane production indicate inhibition of the methane forming bacteria.This suggests that free hydrogen sulphide was not the only compound which inhibited themethanogens.At 55 °C, the sulphate reduction, sulphide concentration and TOC removalefficiency were considerably lower than the mesophilic experiments. This low efficacy canbe attributed to a greater sensitivity of the methanogens, lower inherent activity, or alower concentration of biomass which were active under thermophilic conditions.123Chapter 8Sulphur Management Strategy 2:Inhibition of the Sulphur Reducing Bacteria8.1 OverviewSection 2 explains why inhibition of the SRB by molybdate addition may beof benefit to the anaerobic treatment of sulphur rich effluents.Section 3 describes the results of the experiments which added molybdate tothe effluent at 35 °C.Section 4 describes the results of the experiments which supplemented iron withmolybdate in the effluent at 35 °C.Section 5 describes the results of the experiments which added molybdate tothe effluent at 55 °C.Section 6 summarizes the conclusions from these experiments.8.2 BackgroundSince the oxidized inorganic sulphur species are relatively benign towards theMPB, rather than allowing the oxidized sulphur species to be biologically reduced tosulphide and then removing the sulphide, preventing the SRB from functioning altogetherwould arrest the cause rather than cure the symptom. Khan and Trottier (1978) reportedthe toxic effects of the sulphur compounds on the methanogens increased in the followingorder: sulphate, thiosuiphate, sulphite and sulphide.Inhibiting sulphate reduction to suiphide should prevent the adverse effect ofhydrogen sulphide on methane formation. Saleh et at. (1964) were one of the earlyinvestigation teams who saw the potential of this approach and who catalogued atremendous number of prospective inhibitors. Citing the criteria of economy, ease andsafety of handling, and a lack of reactivity with other materials or bacteria, Saleh et at.124(1964), however, did not screen for molybdate (Mo042jwhich has since been widelycited in the literature as an effective and selective inhibitor of the SRB (Lovley et a!.,1982, Lovley and Kiug, 1983, Norqvist and Roffey, 1983, Winfrey and Ward, 1983,Postgate, 1984, Phelps eta!., 1985, Ueki etaL, 1986 and Jopson eta!., 1987)A mechanism for molybdate inhibition of the SRB has been postulated by someinvestigators. Taylor and Oremland (1987) proposed that molybdate altered the cellularATP balance of the SRB. The basis of molybdate inhibition of the SRB was thought byNorqvist and Roffey (1983) to be a sulphate structure analogue. Newport and Nedwell(1988) noted the stereochemical similarity of molybdate with sulphate and they suggestedthat molybdate may inhibit sulphate transport by the SRB.The range of molybdate concentrations which have been used to inhibit the SRBmay also inhibit the methanogens. Winfrey and Ward (1981) and Puhakka et a!. (1990)observed 20 mM Mo042 to nearly entirely inhibit sulphate reduction. However, Puhakkaet aL (1990) found 20 mM molybdate in both synthetic and NSSC effluents to bebacteriocidic to SRB and bacteriostatic to the MPB. At this molybdate concentration, nomethane was produced from acetate. The concentration at which molybdate inhibits theMPB seems to depend upon the MPB themselves. While 20 mM molybdate inhibitedmethanogenesis in freshwater sediments, it stimulated methane production in marinesediments (Banat etaL, 1981, Sorensen eta!., 1981, Winfrey and Ward, 1981).In an anaerobic filter, Hilton and Archer (1988) found 10 mM to decrease theacetoclastic methane production, resulting in accumulations of acetate. At 5 mMmolybdate, Lovley et a!. (1982) observed inhibition of sulphate reduction to result inmethane production comparable to the rate found in zero sulphate control runs. Phelps eta!. (1985) also added sodium molybdate at 5.0 mM to a MethanosarcinabarkerilDesulfovibrio vulgaris coculture. They measured increased methane production,decreased suiphide formation and a 50% decrease in the SRB mass.125Jopson et a!. (1986) added 2 mM Mo042 with only marginal SRB inhibition atthis level. Although Wu et a!. (1991) found 2.0 mM molybdate to inhibit propionate andethanol conversions by 97 and 29% respectively, they found only slight reductions in theacetate conversion rate at 0.1 to 2.0 mM molybdate in the absence of sulphate. Completeinhibition of the SRB in lake sediments was found to occur at 02 mM molybdate by Smithand Klug (1981), but 20 mM molybdate was required to inhibit the MPB for the samesediment. Even at 1.0 mM molybdate, Puhakka et al. (1990) observed some inhibition ofacidogenesis from glucose as well as acetoclastic methanogenesis. They witnessed a threeweek delay in the onset of methane production from spent neutral suiphite liquor at 1.0mM molybdate.Adding small concentrations of molybdate to anaerobic treatment systems hadsome common effects.a) It lowered the sulphate reduction rate, leading to decreased sulphide levels in both thedigester liquor and the off gas (Norqvist and Roffey, 1983, Winfrey and Ward, 1983,Phelps et a!., 1985, Ueki et a!., 1986, Jopson et aL, 1987). This resulted in decreaseddownstream suiphide laden effluent handling difficulties of oxygen demand, odour, toxicityand corrosion.b) Molybdate addition stimulated hydrogen and methane production by alleviating theSRB competition for hydrogen and acetate (Lovley eta!., 1982, Winfrey and Ward, 1983,Phelps eta!., 1985).The disadvantages of inhibiting the sulphate reducing bacteria by molybdateaddition to the feed include:a) a high, on-going chemical cost since molybdate is a bacteriostatic rather than abactericidal compound (Norqvist and Roffey, 1983, Jopson eta!., 1987),b) the release of molybdate into receiving waters,c) unreliable performance at full scale (Jopson et a!., 1986),126d) inhibition of the MPB at elevated molybdate concentrations (Winfrey and Ward, 1983,Puhakka eta!., 1990), ande) with molybdate inhibition of SRB, it follows that the SRB do not contribute to thetreatment efficiency.8.3 Results: Effect of Molybdate Addition to the Effluent at 35 °Ca) SulphateAt 1.0 mM (242 mg/l), the addition of molybdate to the feed tank at 35 °C resultedin reduced levels of sulphate reduction in all cases. Sulphate reduction decreased markedlyat 1.0 mM (242 mg/i) molybdate (Na2MoO4HO)addition at 35 °C for two out of threewastewater batches (see Table 18). It should be noted that the sulphate removal efficiencyfor the methane phase reactors is an overall efficiency and it includes the sulphate whichwas removed in the acid phase reactors. The results presented below are the quasi-steadystate values and there was no indication from the parameters measured of any adaptationof the SRB towards the molybdate at the levels tested.Longer HRTs promoted sulphate reduction whereas molybdate addition tended toinhibit sulphate removal. This is evident by examining the simple linear regression model:%SO4pp 55...45 *Mo + 36 *HRTAPRr=O.80 p=O.000%SO4rMJJR 66-48 *Mo + 25 *JJRTTr=O.77 p=O.000The control run of no moiybdate addition resulted in sulphate reduction efficienciesof up to 90%. 1.0 mM molybdate additions led to overall sulphate reductions of 15 to40% for two of the three wastewater batches. Three 1.0 mM molybdate experiments wereperformed using three different wastewater batches obtained from the Quesnel River PulpCompany in order to ascertain that this level was indeed the threshold concentration for127inhibition of the SRB. While the mill was reportedly operating on a consistent basis, thebatch to batch variation in the effluent characteristics was considerable. Molybdateaddition at 1.0 mM to the second effluent batch did not confirm the positive findings ofthe first batch. Consequently a third wastewater batch was tested which confirmed theoriginal observation of a significantdecrease-of sulphate reduction. No measured aspect ofthe wastewater characteristics could account for this observation.Table 18: Effect of Molybdate Addition on Sulphate Reduction at 35 °CMean Sulphate Reduction Efficiency j%1 ± CVFeed Acid/Methane Phase Reactor HRT LdayslMolybdate S042±SD APR MPR APR MPR APR MPRAddition [mM] 1mg/I] 0.2 0.4 0.4 0.8 0.6 1.2Obatchi 1533±75 35± 16 59± 12 56± 17 90±23 58±27 89± 110.lbatchl 1528±145 57± 15 87± 11 77±35 87± 11 88± 12 88±60.5batchl 1567±180 49±8 57± 11 66± 18 91±24 56± 13 93± 180.75batchl 1409±70 22±7 33±9 51± 14 81±20 44± 11 82±311.Obatchl 1528±49 17±4 26±7 22± 10 40± 14 15±5 40± 131.Obatch2 1250±62 2±5 9±7 44± 11 75±27 28±6 78± 181.Obatch3 1363±63 8±6 15±3 17±12 23±11 13±7 26±5The 0.1 mlvi molybdate concentration appeared to, in fact, stimulate sulphatereduction. This may be due to molybdate satisfying a nutrient requirement of the SRB orof the MPB which would stimulate acetate uptake and thus alleviate, to some degree, end-product inhibition. Such a general phenomenon is commonly encountered in the study ofinhibitory compounds where a concentration of 1 to 10% of the threshold inhibitoryconcentration stimulates the same microorganisms (Parkin and Speece, 1984). Sulphate128reduction increased with increasing retention time and decreased with increasingmolybdate concentration. Beyond the 0.1 mM concentration, molybdate in the acid phasereactors inhibited sulphate reduction approximately linearly with molybdate concentrationfor the levels tested.b) Total Dissolved SulphideThe inhibition of sulphate reduction by molybdate addition was mirrored inlowered suiphide levels (see Table 19). Similar to the pattern of sulphate reductionefficiency, the total dissolved suiphide concentration in the methane phase reactorsdecreased with an increased concentration of the molybdate added to the feedstock andwith lower HRT.Table 19: Effect of Molybdate Addition on Total Dissolved Suiphide Concentrationat 35 °CMean Total Dissolved Suiphide Concentration ± SD [mg/i]Feed Acid/Methane Phase Reactor HRT [days]Molybdate S2± SD APR MPR APR MPR APR MPRAddition [mM] [mg/I] 0.2 0.4 0.4 0.8 0.6 1.2Obatchi 1±0.1 38±25 67±34 62±28 95±25 68±29 97±320.lbatchl 1±0.1 76±5 104±5 93± 11 120±8 97±8 127±70.5batchl 1±0.3 56±5 69±8 85±9 111±11 75±7 110±70.75batchl 3±0.2 45±8 61±11 91±3 105± 13 76± 10 113±71.Obatchl 1±0.2 4±3 11±5 20±13 30±16 9±3 27±51.Obatch2 3±0.1 6±2 14±4 74±4 79±7 51±4 80±61.Obatch3 2±0.2 29±5 39± 10 70±7 85±8 59±8 102± 12129The simple linear regression models which describe the effects ofmolybdate and HRT on the total dissolved suiphide levels are as follows:S2pR 59 35*Jyfr + 59*HRTApRr=O.59 p=O.O.2O= 7449 *Jyf + 35 *HRTT - -r=O.71 p=O.OO2The sulphide concentration was not as sensitive to molybdate levels as was thesulphate concentration but was consistent with the pattern of sulphate reduction. Whereasno molybdate resulted in total dissolved concentrations of 67 to 97 mg/i in the methanephase reactors, 1.0 mM molybdate led to just 11 to 27 mg/l sulphide for the firstwastewater batch.Molybdate acted in a bacteriostatic, not bactericidal, manner. This phenomenon isdepicted in Figure 5. These graphs plot the recovery of sulphate and sulphideconcentrations in the 10 day period following molybdate addition to the feed at the 1,0mM level. As with the work of Norqvist and Roffey (1983), molybdate in theseexperiments only temporarily inhibited the SRB. Discontinuing the molybdate addition tothe feed tank resulted in a rapid return to the levels of suiphide and sulphate which wereobserved prior to this set of experiments. Ten days after the last batch of molybdatecontaining effluent was processed through the reactors, the sulphate levels which exitedthe MPR returned to the 100 to 200 mg/l sulphate residual. This demonstrated that theSRB were inhibited only as long as molybdate was present in the system.c) Suiphite and ThiosuiphateMost of the samples were spoiled before sulphite and thiosuiphate testing could beperformed. Consequently, a sulphur balance was not performed. However, from thescattering of remaining data for these experiments, a few observations can be reported:130zCl)Cl)C(ID.cf)0HRecovery Time [days] Following Molybdenum AdditionFigure 5: System Recovery Following 1.0 mM Molybdate Addition to the FeedstockRecovery Time [days] Following Molybdenum Addition0 1 2 3 4 5 6 7 8 9 10—.— FeedO 0.2dHRTA.PR1O.4dHRTAPR20— 0.6dHRTAPR3——O.4 dHRTMPR 1—A—08 dHRTMPR2• 1.2dHRTMPR31311. The feed suiphite levels appear to be extremely small, on average 22 mgIl sulphite.2. Sulphite levels were decreased to below detectable limits at all HRTs operated.d) Wastewater Treatment EfficiencyLow levels of molybdate addition resulted in a net stimulation of the organicsremoval capacity, perhaps by filling a metabolic requirement of the MPB or the SRB. At0.1 mM molybdate, the higher sulphate reduction efficiencies and higher concentrations ofsulphide are consistent with stimulated SRBs. These results are presented in Table 20.Table 20: Effect of Molybdate Addition on TOC Removal at 35 °CMean TOC Removal Efficiency j%] ± 5 % CVFeed Acid/Methane Phase Reactor HRT [days]Molybdate TOC APR MPR APR MPR APR MPRAddition [mM] jmgll] 0.2 0.4 0.4 0.8 0.6 1.2Obatchi 2754 22 29 22 25 27 340,lbatchl 2745 21 24 33 35 32 340.5batchl 2735 31 38 33 41 33 420.75 batch 1 2635 19 33 29 38 34 391.Obatchl 2395 22 22 15 36 24 291.Obatch2 2124 13 21 19 22 26 271.Obatch3 2021 19 20 22 22 23 34The TOC removal efficiency for the methane phase reactors is an overall efficiencyand it includes the TOC removed in the acid phase reactors. The major portion of theTOC was removed in the acid phase reactors. The TOC removal generally increased withincreasing URT. The TOC removal efficiency increased from 25 to 34% for the controlruns to a maximum of 42% at 0.5 mM molybdate. Molybdate additions greater than 0.5132mM resulted in reduced treatment efficiencies, ranging from 20 to 36% at 1.0 mM,depending on the wastewater batch. The r2 of 0.38 reflects the absence of linearity ofTOC removal with respect to molybdate and APR HRT.e) Gas ProductionThe mean gas production rates followed the trends similar to the TOC removals(see Table 21). The lowered gas production rates at molybdate concentrations greater than0.1 mM support the conjecture of Puhakka et a!. (1990) that elevated concentrations ofmolybdate inhibited the MPB. Therefore, while low concentrations of molybdate arereasonably effective in inhibiting sulphate reduction, the MPB also appear to be inhibited.The 1.0 mM molybdate addition to effluent batch 1 demonstrated unstable gas productionrates in spite of stable sulphate and TOC removals. Consequently, an average gasproduction rate for this setting was meaningless. From the standpoint of TOC removalefficiency, the optimum concentration of molybdate addition to this effluent approximates0.5 mM. However, this addition does not significantly change the sulphate removalefficiency compared to the runs without molybdate addition.As for the other experiments, the small gas production rates did not account forthe TOC removal (see the carbon balance in Appendix 1). Total gas production increasedwith increasing HRT and decreased with increased molybdate addition above 0.1 mM.These general trends are consistent with TOC removal. A decreased gas production ratewith increased molybdate addition was evidence of molybdate inhibition of the MPB.Lovley and Klug (1983) conversely found that 1.0 to 2.0 mM molybdate inhibited sulphatereduction and stimulated methane formation by alleviating suiphide toxicity and the SRBcompetition for hydrogen.133Table 21: Effect of Molybdate Addition on Total Gas Production Rate at 35 °CMean Gas Production Rate ± SD [I/day]Molybdate Acid+Methane Phase Reactor HRT [days]Addition [mM] 0.6 1.2 1.8Obatchl 0.9±0.2 1.5±0.2 1.2±0.10.lbatchl 1.1±0.2 2.1±0.3 2.6±0.40.5batchl 0.9±0.1 1.6±0.1 2.4±0.40.75 batch 1 0.5 ± 0.2 0.8 ± 0.1 1.3 ± 0.21.Obatchl - - -1.Obatch2 0.3±0.1 0.6±0.1 1.0±0.11.Obatch3 0.1±0.02 0.7±0.1 0.8±0.11) Volatile Fatty AcidsThe mean individual volatile fatty acid concentrations are plotted againstmolybdate addition for the three HRTs of both the APRs and the MPRs. These plots arepresented in Figure 6. An explanation of the trends follows.Formic Acid: Formate concentrations, while highly variable in the feed, were greatlyreduced in the acid phase reactors with ftirther reduction to zero occurring in the methanephase reactors. As the molybdate concentrations increased to 0.75 and 1.0 mM, the formicacid concentration in the acid phase reactors was significantly greater than 0, indicativeperhaps of some degree of MPB inhibition.Linear regression analysis demonstrated that increased molybdate levels led tohigher formate concentrations and a longer APR HRT promoted formate uptake. Thisappears to support molybdate inhibition of formate metabolism by the MPB. The simpleregression model is as follows:FormateAjJR = 2] + 15*Mo - 54*HR TAPRr=O.61 p=O.Ol4134250- 2000200 / ooI::: I 1:::-‘- 50•-0 I I0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00Molybdenum Addition [mM] Molybdenum Addition [mM]500- 2500.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00Molybdenum Addition [mM] Molybdenum Addition [mM]250200- —R— FeedE 0 O.2dHRTAPR1150-. &- O4dHRTAPR2100 -- O.6dHRTAPR3•0.4 d HRT MPR I50—A---O8dHRTMPR2I I • 1.2dHRTMPR30.00 0.25 0.50 0.75 1.00Molybdenum Addition [mM]Figure 6: Effect of Molybdate Addition at 35 °C on Volatile Fatty Acids135Acetic Acid: Similar to the other sulphur management experiments, acetate was producedin the acid phase reactors. A small net consumption of acetate was demonstrated in themethane phase reactors but acetate concentrations in the methane phase reactor exitstream were always large, in the order of 1000 mg/i. The large fluctuations in the feedacetate levels hide any influence that the molybdate level may exerton acetate. This isreflected in the simple linear regression equation which describes the effect of molybdateand HRT on the acetic acid concentration in the MPR:AcetateMpR = 1130 - 440 *Mo + 119*HRTAFRr=0.68 p=O.OO4With the SRB increasingly inhibited, one would expect the acetate producingnature of the SRB to also be inhibited and thus the acetate concentrations from this sourceshould decrease. However, there are other routes by which acetate can be produced.Acidogenic and acetogenic non-SRB produce acetate and may fill a niche previouslyoccupied by the SRB. Puhakka et a!. (1990) have claimed that the molybdate levels whichinhibit the SRB also begin to inhibit the MPB from utilizing acetate.Propionic Acid: Similar to the other VFAs in this set of experiments, the propionic acidlevels in the feedstock varied significantly from batch to batch. At low molybdate levels,the propionic acid concentrations in the feed were largely catabolized in the acid phasereactors by the SRB and non-SRB acidogens. For these experiments, the propionateconcentrations in the methane phase reactors were not significantly different from those ofthe acid phase reactors. Since the MPB were only marginally active and there was littleactivity of the SRB in the methane phase reactors, hydrogen uptake would also be slow,thereby decreasing the thermodynamic attractiveness of oxidizing propionate to hydrogenand acetate. Increased molybdate levels to 1.0 mM resulted in the rate of propionategeneration surpassing the rate of propionate consumption. This was expected since theSRB, consumers of propionic acid, were inhibited.136Butvric Acid: The concentration of n-butyric acid in the feed was generally greater thanthat measured in the reactors. It appears that n-butyrate was consumed in the acid phasereactors and was generated in the methane phase reactors at low levels of molybdate. Asthe molybdate concentration increased above 0.5 mM, this trend was reversed. The acidphase reactor butyrate concentration- decreased successively in the acid phase reactor andalso in the methane phase reactor.Lactic Acid: Lactate was consumed in both the acid phase reactors and the methanephase reactors. Significant lactate concentrations still remained in the exit stream of themethane phase reactors. In spite of its status as a preferred substrate of the SRB, no trendof lactate concentration with molybdate addition was observed. Even if completeinhibition of the SRB was achieved and the SRB did not consume lactate, the acidogenicmicroorganisms could still degrade lactate to propionate or acetate as long as thehydrogen concentrations were kept low.g) Reactor pHThe pH of the acid phase reactors was targeted at 7.5, the setting whichdemonstrated the best results in the previous acid phase reactor experiments. Themeasured methane phase reactor pH levels were lower (6.8 to 7.5) in every case than theacid phase reactors as VFAs were generated but not utilized to an appreciable extent bythe microorganisms. These results are summarized in Table 22.h) Sodium Carbonate ConsumptionIn order to maintain a stable pH of 7.5 in the acid phase reactors, the sodiumcarbonate consumption generally increased with increased HRT (see Table 23). Thisindicates the slow rates of acidogenesis in these reactors. The sodium carbonateconsumption rate decreased with the addition of molybdate to the feedstock. Theregression model showed only a weak linear association:137Table 22: Effect of Molybdate Addition on Reactor pH at 35°CMean Reactor pH ± SDFeed Acid/Methane Phase Reactor HRT Ldays]Molybdate pH ± APR MPR APR MPR APR MPRSDAddition jmM] 0.2 0.4 0.4 0.8 0.6 1.20 batch 1 7.0±.2 7.6±.1 6.8±.4 7.6±.1 7.1±.2 7.5±.1 7.1±.10.lbatchl 6.9±.2 7.5±.1 7.2±.1 7.5±.1 7.4±.1 7.5±.1 7.4±.10.5 batch 1 6.7±.2 7.5±.1 7.0±.1 7.5±.1 7.4±.1 7.5±.1 7.5±10.75 batch 1 5.9±.2 7.5±.1 7.0±.1 7.5±1 7.5±.1 7.5±.1 7.4±.11.0 batch 1 7.0±.1 7.6±.1 6.9±.3 7.6±.1 7.2±.2 7.5±.1 7.2±.11.Obatch2 6.0±.4 7.5±.1 6.8±.1 7.5±.1 7.4±.1 7.5±.1 7.3±.11.Obatch3 7.7±.1 7.7±.1 7.3±.1 7.5±.1 7.5±.1 7.5±.1 7.4±.1gNa2CO3/lfeed= 1.6.O.6*Mo + 1.2*HRTApRr=O.55 p=O.O39There are a number of possible explanations for this observation. Inhibiting theSRB decreased their ability to both generate and utilize the propionic, butyric and lacticacids. If the acidogenic SRB were inhibited, they would also produce smaller amounts ofVFAs. Lower sulphide concentrations may promote the activity of the MPB which wouldtherefore consume greater amounts of VFAs. The accumulations of such acids wouldcause the pH to decrease and sodium carbonate consumption to rise to maintain the pHsetpoint.138Table 23: Effect of Molybdate Addition on Carbonate Consumption at 35 °CMean Sodium Carbonate Consumption LgIl feed]Molybdate Acid Phase Reactor HRT [days]Addition [mM] 0.2 0.4 0.6Obatchl 1.76 1.81 2.230.lbatchl 1.70 1.88 2.100.5batchl 1.36 1.64 1.670.75 batch 1 1.57 2.47 2.311.Obatchl 1.49 2.19 2.231.Obatch2 0.70 1.20 1.041.Obatch3 0.28 0.73 0.658.4 Results: Effect of Iron and Molybdate Addition at 35 °CMolybdate addition to the feedstock, even at 1.0 mM levels, did not entirelyeliminate sulphate reduction to suiphide (see Table 18). High suiphide levels persistedwhich may have exerted inhibitory effects on the waste treatment microbes.Augmenting the SRB inhibition strategy with suiphide removal via precipitation asiron suiphide was performed for this set of experiments. This was done in order toexamine the effects of high sulphate and low sulphide levels on the wastewater treatmentmicroorganisms. Significant variability in the feedstock between experiments wasminimized for these runs since they were performed using the same effluent shipment.Adding an excess of iron to the wastewater would have additional benefits besidesjust the chemical precipitation of suiphide. DTPA would be complexed and perhaps maketrace metal nutrients available to the microorganisms. Also, sulphate and iron have beenshown by van den Berg et al. (1980) and Callander and Barford (1983) to interact139strongly to stimulate the conversion of acetate to methane. Laanbroek et aL (1984)observed Desu41fovibrio baculatus to be dependent upon the availability of sufficient iron.a) Sulphate1.0 mM molybdate with 5 mg/l FeC136H2Osignificantly decreased the sulphatereduction efficiency (Table 24). It should be noted that the sulphate removal efficiency forthe methane phase reactors is an overall efficiency and it includes the sulphate which wasremoved in the acid phase reactors. At the nominal 1.2 and 1.8 day HRTs, the totalsulphate reduction was decreased from 87 and 85% to 23 and 26% respectively. Ironlevels at 100 mg/I FeC136H2Oplus 1.0 mM molybdate added to the feedstock appearedto only slightly increase the reduction of sulphate compared to the 5 mg/i FeC136H2Oplus 1.0 mM molybdate. In the 200 mg/i ferric chloride plus 1.0 mM molybdateexperimental run, the sulphate reduction efficiency was significantly increased, especiallyat the intermediate and longest HRTs.It is postulated that the iron could have played one or more of the following roles:1) to act as a precipitating agent and remove sulphide, or2) to act as an essential nutrient for the microorganisms (Laanbroek et aL, 1984, Isa et a!.,1986), or3) to interfere with the inhibitory effects of the organic compounds or of molybdate on theSRB, or4) to interact with the DTPA chelating agent to displace and liberate other chelated micronutrients, thereby making them available to the SRB.b) Total Dissolved SuiphideIn the absence of molybdate and just 5 mg/l FeC136H2O (the baseline ironaddition for all of the experiments), the total dissolved sulphide concentrations in thereactors increased up to 266 mg/l at the longest HRT for this effluent (see Table 25). This140was the highest suiphide concentration which was measured in any of the experiments andwas in the range (greater than 200 mg/i) which has been suggested in the literature to bestrongly inhibitoiy towards the MPB.Table 24: Effect of Molybdate and Iron Addition on Sulphate Reduction at 35 °CMean Sulphate Reduction Efficiency [%] ± CVFeed Acid/Methane Phase Reactor HRT [days]Chemical S042±SD APR MPR APR MPR APR MPRAddition [mg/I] 0.2 0.4 0.4 0.8 0.6 1.20mM Mo+5mg/1FeC136H2O 1369±13 7±8 30± 10 39± 11 87± 18 28±5 85±6(control)1.0 mMMo+5mg/1FeC136H2O 1363±63 8±6 15±3 17± 12 23± 11 13±7 26±51.0mMMo+ 100mg/iFeC136H2O 1114±16 3±10 11±7 12±8 33± 15 11± 10 39± 151.0 mMMo+200 mg/IFeC136H2O 1338±15 22±9 22± 10 56±22 87± 19 47± 10 100±0The addition of 1.0 mM sodium molybdate (with 5 mg/i ferric chloride) to thiswastewater decreased the extent of sulphate reduction but still significant suiphideconcentrations in the liquor were measured. 100 mg/i ofFeC136H2Oadded with 1.0 mMmolybdate had no significant effect on the dissolved suiphide levels for this effluent.Paradoxically but consistent with the measured sulphate concentrations, ferric chlorideaddition at the 200 mg/i level with 1.0 mM moiybdate increased the dissolved suiphideconcentration almost to the level which was observed without molybdate or iron. Theprobable explanations for these results -- iron acting as a precipitating agent, as a nutrient,141as an antagonist or as a sacrificial cation -- are outlined under the sulphate discussion andalso apply to sulphide.Table 25: Effect of Molybdate and Iron Addition on Total Dissolved SulphideConcentration at 35°CMean Total Dissolved Suiphide Concentration ± SD 1mg/i] -Feed Acid/Methane Phase Reactor HRT [days]Chemical 2- SD APR MPR APR MPR APR MPRAddition jmg/l] 0.2 0.4 0.4 0.8 0.6 1.20mM Mo+5mg/1FeC136H2O 3.1±1.0 132±11 186±3 211±6 264±3- 170±7 266±6(control)1.0 mMMo± 5 mg/l FeC136H2O 2.0±0.2 29±5 39±10 70±7 85±8 59±8 102±121.OmMMo+ 100 mg/lFeC136H2 2.8±1.2 27±3 47±8 51±5 89± 17 50±7 109± 141.OmMMo+200mg/lFeC136H2 2.8±1.1 98±9 139± 11 186±19 213± 14 147±20 232± 15c) Suiphite and ThiosuiphateIncreased iron additions in the presence of 1.0 mM molybdate had the effect ofincreasing the concentrations of both sulphite and thiosuiphate (see Tables 26 and 27). Forsome of the reactors, the concentration of suiphite was greater than in the incoming feed,suggesting that sulphite was an intermediate product in the sulphur reducing cycle.Trudinger (1981) claimed that the pathway of the reduction of suiphite to sulphide wasunknown. Inconsistent with these results are the reports of some investigators (Puhakka eta!., 1984, Pipyn et aL, 1985) who have claimed that suiphite was not reduced to suiphideunder anaerobic conditions and that suiphite demonstrated toxic effects at 100 mg/i.According to Nakatsukasa and Akagi (1969), Haschke and Campbell (1971),Hatchikian (1974), the reduction of thiosulphate to sulphide proceeds via sulphite. The142trends of thiosuiphate concentration were similar to those of suiphite but were moremarked for the thiosuiphate assays. The thiosuiphate concentrations increased withincreasing HRT and were significantly greater than in the feed.Table 26: Effect of Molybdate and Iron Addition on Suiphite Concentrationat 35 °CMean Suiphite Concentration ± SD [mg/i]Feed Acid/Methane Phase Reactor HRT [days]Chemical so32- APR MPR APR MPR APR MPR±SDAddition [mg/I] 0.2 0.4 0.4 0.8 0.6 1.20mM Mo+5mg/1FeC136H2O 22±4 0±0 5±4 3±3 28±18 0±0 21±21(control)1.0 mMMo+lOOmg/lFeC136H2 45±36 13±5 16±4 34±3 32±5 29±4 23±61.OmMMo+ 200 mg/l FeC136H2O 27±4 22 ± 7 43 ± 7 16 ± 3 45 ± 4 20 ± 8 22 ± 8Table 27: Effect of Molybdate and Iron Addition on Thiosuiphate Concentrationat 35 °CMean Thiosulphate Concentration ± SD [mg/I]Feed Acid/Methane Phase Reactor HRT [days]Chemical S203 APR MPR APR MPR APR MPR±SDAddition [mg/I] 0.2 0.4 0.4 0.8 0.6 1.20mM Mo+ 5 mg/i FeC136H2O 65±7 57±6 69±8 66±7 103±3 57±2 89±6(control)1.OmMMo+ 100mg/iFeC136H2 65±16 126±27 84±8 182±31 160±34 161±26 141±281.0 mMMo+200 mg/iFeC136H2O 48±18 151±14 165±18 130±7 156±14 162±13 174±10143d) Wastewater Treatment EfficiencyIron chloride which was added at 100 mg/i plus 1.0 mM molybdate generallyimproved the treatment efficiency for all but the highest HRT over that found in the 5 mg/IFeCl36H2Owithout molybdate control, or with the 1.0 mM molybdate experiments (seeTable 28). The TOC removal efficiency for the methane phase reactors is an overallefficiency and it includes the TOC removed in the acid phase reactors. The gas productionrate was smaller at this setting compared to both the control and the 200 mg/i ferricchloride plus 1.0 mM molybdate addition. Lowered gas production does not hint at adecreased inhibition of the MPB.This improved treatment efficiency with iron addition may have been due to anumber of causes.1) Precipitated suiphide led to lower inhibition of the waste treatment microorganisms,since the highest TOC removals coincided with lowered sulphate removal efficiencies andsulphide concentrations less than the commonly quoted 200 mg/i level where inhibitionbecomes significant (see Tables 24 and 25).2) The SRB were stimulated, although enhanced sulphate reduction was not correlatedwith greater TOC removal.3) Organic compounds were precipitated by iron, since the enhanced TOC removal wasnot accounted for by an increase in gas production.4) DTPA was complexed by iron. Kennedy et al. (1991b) described the function of suchiron addition as that of a sacrificial trivalent cation. They postulated that ferric chlorideaddition would cause the DTPA to preferentially bind the ferric chloride and thus liberateother micronutrients, making them biologically available to the waste treatment microbes.This theory was not tested for these experiments.As the iron concentration was increased from 100 to 200 mg/l, the treatmentefficiency decreased back to the levels measured in the control (0 mM molybdate, 5 mg/Iiron chloride). At the larger HRTs, the TOC concentrations appeared to be unaffected. In144particular, the TOC removal efficiency in the APRs at the two lowest HRTs was negligiblefor the 200 mg/l ferric chloride experiments. This is perhaps indicative of 200 mgfl offerric chloride exceeding the threshold inhibitory concentration so that the chemicalprecipitation of organic compounds roughly compensated for the compromised biologicalreduction.Table 28: Effect of Molybdate and Iron Addition on TOC Removal at 35 °CMean TOC Removal Efficiency [%] ± 5 % CVFeed Acid/Methane Phase Reactor HRT Idays]Chemical TOC APR MPR APR MPR APR MPRAddition [mg/i] 0.2 0.4 0.4 0.8 0.6 1.20mM Mo+ 5 mg/i FeC136H2O 2690 5 22 22 24 17 24(control)1.0 mM Mo+ 5 mg/i FeC136H2O 2020 19 20 22 22 23 341.0 mM Mo+ 100 mgfFeCl36H2O 1965 25 26 27 32 24 311.OmMMo+ 200 mg/i FeC136H2O 2575 0 22 1 24 17 24e) Gas ProductionIron addition, at either 5, 100 or 200 mg/l, had no statistically significant effect onthe measured gas production. Adding molybdate at 1.0 mM reduced the gas productioncompared to the control setting. Consistent with TOC removal patterns, gas productionincreased with increasing HRT. Table 29 summarizes these results.1451) Reactor pH and Sodium Carbonate ConsumptionIron plus molybdate addition to the feedstock did not influence the pH of themethane phase reactors (see Table 30). However, sodium carbonate consumption in theacid phase reactors increased substantially with increased iron addition (Table 31) in orderto counter the acidity and to maintain an acid phase reactor pH of 7.5. Even at the highestsodium carbonate consumption rates in Table 31, dilution of the reactor liquor by thesodium carbonate solution was less than 5%.Table 29: Effect of Molybdate and Iron Addition on Total Gas Production Rateat 35 °CMean Gas Production Rate ± SD [l/dayjChemical Acid+Methane Phase Reactor HRT [daysiAddition 0.6 1.2 1.80mM Mo+ 5 mg/i FeC136H2O 0.32±04 0.64±02 1.14±.06(control)1.OmMMo+ 5 mg/I FeC136H2O 0.11±.02 0.72±. 10 0.76±.061,0 mMMo+ 100 mg/i FeC136H2O 0.19±.06 0.65±19 0.66±. 121.OmMMo+ 200 mg/I FeC136H2O 0.25±.12 0.61±. 12 0.90±.33146Table 30: Effect of Molybdate and Iron Addition on Reactor pH at 35 °CMean Reactor pH ± SDFeed Acid/Methane Phase Reactor HRT [days]Chemical pH ± APR MPR APR MPR APR MPRSDAddition 0.2 0.4 0.4 0.8 0.6 1.20 mM Mo,+ 5 mg/I 7.5±.2 7.6±.l 7.2±.18 7.5±.1 7.5±.l 7.5±.1 7.4±.1FeCI3 6H20(control)1.0 mMMo+ 5 mg/i 7.7±.l 7.7±.l 7.3±.11 7.5±.1 7.5±.l 7.5±.1 7.4±.1FeC136H2O1.0 mMMo+ 100 mg/I 7.8±.1 7.6±.l 7.3±.07 7.6±.l 7.5±.1 7.5±.l 7.4±.lFeC136H2O1.0 mMMo+200mg/I 7.8±.l 7.5±.l 7.3±.l0 7.5±.1 7.6±.1 7.5±.l 7.4±.lFeC11 6H,OTable 31: Effect of Molybdate and Iron Addition on Carbonate Consumptionat 35°CMean Sodium Carbonate Consumption [gil feediChemical Acid+Methane Phase Reactor HRT [days]Addition [mM] 0.2 0.4 0.60 mM Mo,+ 5 mg/i FeC!3 6H20 0.27 0.74 0.65(control)1.OmMMobatch3+ 5 mg/! FeC!3 6H20 0.28 0.73 0.651.0 mMMo+ 100 mg/i FeC!3 6H20 0.45 0.80 0.761.0 mMMo+ 200 mg/i FeC136H2O 0.67 1.20 0.961478.5 Results: Effect of Molybdate Addition to the Effluent at 55 °Ca) SulphateSulphate reduction at 55 °C increased with increased HRT and decreased withincreased molybdate addition. These trends are summarized in Table 32. The sulphateremoval efficiency for the methane phase reactors is an overall efficiency and it includesthe sulphate which was removed in the acid phase reactors. Sulphate reduction rangedfrom 40 to 50% for the 0.6 and 1.8 day HRT control runs without molybdate. Molybdateaddition of 0.5 mM and 1.0 mM decreased these reductions to approximately 30%. Theeffect of molybdate addition and HRT on the sulphate reduction efficiency is describedusing simple linear regression analysis:%SO4rApR 35. - 5*HRTAPRr=O.94 p=O.O.2O%SO4rMpR =39-13 *Mo + 5 *HRTMPRr=O.93 p=O.OO.2Such low sulphate reductions may have been due to a low density or activity of theSRB which were started from the mesophilic inocula. Alternatively, higher temperaturesmay have caused the SRB to be more inhibited than at 35 °C. This could have been due toa greater susceptibility of the microorganisms to inhibition at higher temperatures or dueto changes in the physical properties of the inhibiting compounds themselves. Candidatecompounds for such inhibition include: sulphide or the oxidized inorganic sulphurcompounds, resin acids or other wood extractives or DTPA.While the sulphate reduction efficiency decreased markedly at 1.0 mM (242 mg/I)molybdate (Na2MoO4HO)addition at 35 °C, there is no evidence at 55 °C of such aninhibition threshold at levels up to 1.0 mM molybdate. Instead, molybdate addition to thefeedstock inhibited sulphate reduction approximately linearly with molybdateconcentration and decreased with decreasing FIRT.148Table 32: Effect of Molybdate Addition on Sulphate Reduction at 55 °CMean Sulphate Reduction Efficiency j%I ± CVFeed Acid/Methane Phase Reactor HRT [days]Molybdate SO42- APR MPR APR MPR APR MPR±SDAddition [mM] [mg/li 02 0.4 0.4 0.8 0.6 1.20 1410±2 32±7 40±6 36±7 43±9 34±5 50±40.5 1480±7 23±6 34±2 26±6 36±6 20±9 37±31.0 1370±3 19±3 30±5 20±4 33±7 17±5 33±6b) Total Dissolved SuiphideThe mean total dissolved suiphide concentrations followedsulphate reduction efficiencies, decreasing with decreasing HRTmolybdate concentration (Table 33).trends similar to theand with increasingTable 33: Effect of Molybdate Addition on Total Dissolved Sulphide Concentrationat 55°CMean Total Dissolved Sulphide Concentration ± SD [mg/i]Feed Acid/Methane Phase Reactor HRT [days]Molybdate SO42- APR MPR APR MPR APR MPR±SDAddition [mM] [mg/li 0.2 0.4 0.4 0.8 0.6 1.20 1410±28 42±7 61±10 48±1 62±5 50±7 83±90.5 1480±103 25±4 60±4 37±7 63±4 29±8 63±111.0 1370±41 26±5 49±4 31±3 47±5 24±3 55±6The regression models describing the effects of sulphur management and HRT onthe dissolved suiphide concentration gave good fits:149S2ApR =43-20 *Mo + 5 *HRTAPRr=0.86 p=0.020S2MpR — 60- 18*Mo + JQ*HRTTr=0.89 p=O.Ol0Feed suiphide concentrations werelow, 2to 4 mg/i. The reactor dissolved suiphideconcentrations, while problematic from a downstream effluent point of view, should exertno significant inhibitory effects on the MPB according to the observations of mostauthors.As for the mesophilic experiments, molybdate acted in a bacteriostatic, notbactericidal, manner, only temporarily inhibiting the SRB. Discontinuing the molybdateaddition to the feed tank resulted in a rapid return to the sulphide levels observed prior tothis set of experiments.c) Suiphite and ThiosuiphateSulphite levels in the feed and reactor samples were very small and demonstratedtrends with respect to HRT or molybdate addition as described by the regression model(also see Table 34):SO3ApR = 3 + 5 *Mo + 2 *HRTAPRr=0.89 p=O.OJOSO3MPR = 4 + 6 *Mo + 0. 6*HRTTr=0.74 p=O.090The measured suiphite concentrations of 9 to 29 mg/i in the feedstock should present noobstacle to the wastewater treatment microorganisms. Suiphite levels were lower in thereactors than in the feedstock, indicating some degree of conversion, at odds with someauthors (Puhakka et al., 1985, Mebrotra et ai., 1987). There appeared to be no differencebetween the acid phase reactor and the methane phase reactor with respect to sulphite.150Table 34: Effect of Molybdate Addition on Suiphite Concentration at 55 °CMean Sulphite Concentration ± SD 1mg/i]Feed Acid/Methane Phase Reactor HRT [days]Molybdate SO32- APR M.PR APR MPR APR MPR±SDAddition [mM] [mg/i] 0.2 0.4 0.4 0.8 0.6 1.20 10±3 4±1 4±1 4±1 7±2 5±2 5±10.5 9±2 4±2 4±1 5±1 6±1 5±1 5±11.0 29±6 9±2 9±2 9±2 15±1 10±1 11±2Thiosuiphate concentrations increased in the reactors compared to the influent. Asummary of these measurements is presented in Table 35. No linear trend is evident fromregression analysis with respect to thiosulphate concentration and molybdate addition orHRT (APR r2 = 0.032, MPR r2 = 0.237).Table 35: Effect of Molybdate Addition on Thiosuiphate Concentration at 55 °CMean Thiosuiphate Concentration ± SD [mg/i]Feed Acid/Methane Phase Reactor HRT [days]Molybdate S203 APR MPR APR MPR APR MPR±SDAddition [mM] [mg/i] 0.2 0.4 0.4 0.8 0.6 1.20 0±0 12±3 47±3 54±5 52±11 72±1 32±30.5 0±0 41±1 36±2 75±16 62±1 62±12 22±51.0 3±1 36±7 45±15 108±13 86+12 83±1 16±7While the feed thiosuiphate concentrations were barely detectable, those in the acidand methane phase reactors were quite significant. Generally, the thiosulphate levels151appeared to increase with increased molybdate concentration. No simple relationship wasrecognized with respect to HRT. Thiosulphate appears to have been generated in the acidphase reactors and reduced in the methane phase reactor.d) Wastewater Treatment EfficiencyTable 36 shows that the TOC removal efficiencies were low (maximum 32%) foreach of the molybdate experiments. It should be noted that the TOC removal efficiency forthe methane phase reactors is an overall efficiency and it includes the TOC which wasremoved in the acid phase reactors. This relative lack of treatment was also reflected in thevery small gas production rates (Table 37). At 0.5 mM molybdate addition to the feedtank, the TOC removal efficiency was unchanged from the zero molybdate control. TOCremoval increased at the 1.0 mM molybdate addition setting from approximately 20% toapproximately 30%.Unlike the 35 °C runs, where the treatment efficiency declined with increasinginhibition of the SRB (Table 18), thermophilic operation saw the higher TOC removalscoinciding with lower sulphate reduction and lower dissolved sulphide concentrations. At55 °C, the inhibitory effects of total dissolved suiphide may be greater than at 35 °C. Thisis consistent with many of the literature findings of increased sensitivity at thermophilictemperatures.The effect of molybdate concentration and HRT on TOC removal efficiencydemonstrated a poor fit for the APR (r2 = 0.46). The model was improved however forthe MPR:%TOCrMpR = 18+ 10 *Mo + 0.6 *HRTTr=0.76 p=O.O8OThis model indicates that there was no significant effect of HRT on the TOCremoval efficiency for these experiments. This indicates that a small portion of the TOC isfairly biodegradable but the remainder is recalcitrant to the microorganisms under these152treatment conditions. Typically, 65 to 90% of the total TOC removal occurred in the acidphase reactors for the 0 and 0.5 mM runs. At 1.0 mM molybdate, 59 to 72% of the TOCwhich was removed was effected in the acid phase reactors. The enhanced TOC removalat 1.0 mM molybdate indicates an alleviation of inhibitory effects or stimulation bymolybdate.Table 36: Effect of Molybdate Addition on TOC Removal at 55 °CMean TOC Removal Efficiency j%] ± 5 % CVFeed Acid/Methane Phase Reactor HRT [days]Molybdate TOC APR MPR APR MPR APR MPRAddition [mM] [mg/I] 0.2 0.4 0.4 0.8 0.6 1.20 2530 13 20 16 18 19 230.5 2425 15 19 19 21 15 171.0 2790 20 29 17 29 23 32e) Gas ProductionSimilar to the other experiments, a considerable portion of the TOC which wasremoved was not accounted for by the gas produced. The low gas production rates appearto be due to inhibition of the acetoclastic MPB. These results are presented in Table 37.Table 37: Effect of Molybdate Addition on Total Gas Production Rate at 55 °CMean Gas Production Rate ± SD EL/day]Molybdate Acid+Methane Phase Reactor HRT [days]Addition [mM] 0.6 1.2 1.80 0.10±.04 0.12±.05 0.30±120.5 0.08±05 0.14±07 0.27±101.0 0.12±.07 0.24±.03 0.36±.12153Increased gas production rates were observed with increased HRT and withincreased molybdate concentrations according to the following equation:gas (i/day) = -0.03 + 0.07*Mo + 0.2*HRTTr=0.93 p=O.0031) Volatile Fatty AcidsGraphs depicting the VFA concentrations for the molybdate experiments at 55 °Care presented in Figure 7. Very brief explanations for the VFA trends follow below.Formic Acid: Formate was reduced to near zero levels for all of the reactors at allmolybdate levels. The marginally higher formic acid concentrations may be reflective ofMPB inhibition at the 1.0 mM molybdate setting. This is indicated in the regression model:FormateMpR — 0.90 - 1.0 *Mo -0. 06*HRTTr=0.87 p=O.Ol5Acetic Acid: Acetic acid levels were increased in the acid phase reactors over the feedconcentrations of approximately 1000 mg/l. The acetate concentrations in the methanephase reactor were even greater than in the APRs. This indicates that acetate wasgenerated but most of it was not consumed, opposite of what could be anticipated. Whenmolybdate inhibited the SRB’s ability to reduce sulphur to suiphide, their ability to produceacetate from the catabolism of long chained fatty acids should also be diminished. Theacetate concentration could be expected to decrease rather than increase, especially if thelowered sulphide levels alleviated MPB inhibition.An increase in the molybdate level to 1.0 mM led to higher acetate concentrations in allreactors. No trend with HRT is apparent.154Molybdenum Addition {mM]Figure 7: Effect of Molybdate Addition at 55 °C on Volatile Fatty Acids400•3002001002500I::::1000•500I I I I0.00 0.25 0.50 075 1.00Molybdenum Addition [mM]0.00 0.25 0.50 0.75 1.00Molybdenum Addition [mM]250E 200C)— 150•iooc-4 50200150•oo•50.0.0 0.5 1.0Molybdenum Addition [mM]0.00 0.25 0.50 0.75 1.00Molybdenum Addition [mM]400•- 300C)C) 200C)1000.0—— FeedO.2dHRTAPR1O.4dHRTAPR20— O.6dHRTAPR3—•—O.4 dHRT MPR 1—A—08 dHRT MPR2O 1.2dRRTMPR30.5 1.0155Propionic Acid: In general, propionic acid levels were greater, in the reactors than in thefeed. The propionic acid concentrations decreased with increasing HRT. The APRconcentrations were diminished somewhat in the MPRs. These effects are described by thelinear model:PropionateApR = 26070*Mo -2i0HR’TApR “ ‘‘ - -r=0.934 p=O.0O2The effects of molybdate addition and FIRT on the propionic levels in the MPRs isnot discernible from these experiments (r2 = 0.39).Butvric Acid: The changing influent n-butyrate concentrations make identification of anytrends in the APRs or the MPRs with either HRT or molybdate addition difficult (APR r2= 0.24 MPR r2 = 0.48). Without molybdate, butyrate was generated in the reactorswhereas at 1.0 mM molybdate, butyrate was oxidized to acetate. The higher acetate levelsat 1.0 mM molybdate are consistent with decreased butyrate concentrations.Lactic Acid: Influent lactic acid levels were roughly halved in the APRs to roughly 80mg/l at all molybdate levels tested. The MPR lactic acid concentrations were only slightlybelow those of the APRs. The residual lactate, propionate and butyrate concentrations areindicative that the SRB were not deficient in a suitable carbon source. Neither themolybdate concentration nor the HRT appear to exert an effect on the lactateconcentration for the levels tested.g) Reactor pHThe methane phase reactor pH declined to near neutral levels following thetargeted acid phase reactor pH of 7.5. This is shown in Table 38. There appears to be nosimple linear relationship between molybdate addition or HRT and the methane phasereactor pH (r2 = 0. 045).156Table 38: Effect of Molybdate Addition on Reactor pH at 55°CMean Reactor pH ± SDMolybdate Feed Acid/Methane Phase Reactor HRT Idays]Addition pH ± APR MPR APR MPR APR MPRSD1mM] 0.2 0.4 0.4 08 0.6 1.20 6.3±.1 7.5±.1 6.9±.3 7.5±.1 7.1±.1 7.6±.1 7.0±.10.5 6.3±.1 7.5±.1 7.1±.1 7.5±.1 7.1±.1 7.6±.1 7.1±.21.0 6.3±.2 7.5±.1 7.O±.2 7.6±.1 7.3±.1 7.5±.1 7.O±.1h) Carbonate ConsumptionTable 39 summarizes the sodium carbonate consumption required in order tomaintain an acid phase reactor pH of 7.5 for the varying additions of molybdate at 55 °C.In comparison with the zero molybdate control run, the mean sodium carbonateconsumption decreased when 1.0 mM molybdate was added to the feed. The 0,5 mMmolybdate level required significantly more carbonate than either the control or the 1.0mM molybdate experiment in order to maintain the acid phase reactor pH at 7.5. This pH7.5 setting coincided with the maximum VFA production. Lower sodium carbonateconsumption at the 1.0 mM molybdate level may be due to lower acid production by theSRB and/or by the non-sulphur reducing acidogens or by increased acid consumption bythe MPB.The mean carbonate consumption increased with increased 1-IRT, indicating thatacidification of the rather recalcitrant effluent was not complete at 0.2 days but rather wasa function of time between 0.2 and 0.6 days.157Table 39: Effect of Molybdate Addition on Carbonate Consumption at 55 °CMean Sodium Carbonate Consumption jgfl feed]Molybdate Acid Phase Reactor HRT LdayslAddition [mM] 0.2 0.4 0.60 0.82 0.90 1.130.5 1.56 1.46 1.761.0 0.53 0.78 0.408.6 Conclusions: Effect of Molybdate AdditionAt 35 °C, molybdate addition at 1.0 mM markedly inhibited sulphate reduction,from 90% to approximately 26% with a corresponding decrease in the dissolved sulphidelevels to non-inhibitory concentrations. This inhibition of the sulphur reducing bacteria didnot result in increased TOC removal. Molybdate acted in a bacteriostatic, not bactericidalmanner, where the inhibition was reversed upon discontinuing molybdate addition. Thehigh acetate concentrations and decreased methane production rates indicate that, at thelevels where molybdate was effective in inhibiting the SRB, the methanogens were alsonegatively affected. Thus molybdate addition to this mechanical pulping effluent does notappear to be a practical sulphur management strategy.Similar to the experiments which varied the acid phase reactor pH, 55 °C treatmentexhibited lower performance. The already low SRB activity in the absence of molybdatemade the effects of molybdate addition at levels up to 1.0 mM difficult to discern. A smallincrease in the TOC removal efficiency was observed at 1.0 mM, suggesting either that themethanogens use molybdate or that the decrease in dissolved sulphide levels relievedinhibition of the MPB.The extremely high acetate concentrations and low methane production rates atboth temperature regimes indicate that the methane producing bacteria were strongly158inhibited. Probable candidates for the inhibitory compounds include: molybdate itselfsulphate, low levels of suiphide, wood extractives or DTPA.159Chapter 9Sulphur Management Strategy 3:Removal of Suiphide from Solution9.1 OverviewSection 2 explains why removing suiphide by means of precipitation or gasstripping may be useful in the anaerobic treatment of sulphur rich effluents.Section 3 describes the results of gas stripping the dissolved suiphide from themethane phase reactors at 35 °C.Section 4 describes the results of gas stripping the dissolved sulphide from themethane phase reactors at 55 °C.Section 5 summarizes the conclusions from these experiments.9.2 BackgroundDissolved suiphide can be removed from the bulk liquor either by precipitation orby stripping. Both techniques are briefly reviewed.Precipitating SuiphideThe method most commonly employed to inactivate the toxic effects of hydrogensulphide is to precipitate the suiphide from solution as an insoluble iron suiphide by addingiron salts to the liquid feed (Winfrey and Zeikus, 1977, Love, 1987, Jensen et al., 1988,Sarner et al., 1988). This precipitation proceeds according to the reaction (Sarner et al.,1988):Fe2 + 112S — FeS + 2HFerric chloride or ferrous sulphate has demonstrated effective removal of dissolvedsulphide. Forming an insoluble iron sulphide complex removes suiphide from the liquorand thus hydrogen sulphide does not contaminate the methane/carbon dioxide gas mixturewhich is formed. It increases the TOC removal and sulphate reduction since both the SRB160and MPB are active without suiphide inhibition (Isa et al., 1986). Although widelypracticed, the technique does suffer from a number of shortcomings.a) Iron salts are expensive chemicals (unless an iron rich waste stream such as waste pickleliquor is available) and considerable quantities are required (Winfrey and Zeikus, 1977,Jopson et al., 1987, Love, 1987, Jensen et aL, 1988, Sarneret a!., 1988, Buismanet aL,1991). About 3 kg of ferric chloride are required to precipitate 1 kg of sulphide (Love,1987).b) Accumulation of FeS precipitate in reactors leads to problems of plugging and apreferential retention of the heavy iron sulphide precipitate over the active biomass(Winfrey and Zeikus, 1977, Callander and Barford, 1983, Jopson et a!., 1987, Love, 1987,Sarner et a!., 1988). Ultimately, these buildups of precipitate will require disposal whichcarries with it extra costs.c) Sulphur precipitation as FeS renders sulphur recovery unfeasible (Love, 1987).d) Ferric chloride is corrosive, adding to the cost of materials for any contactedequipment.e) The blackened effluent which results due to the precipitation as FeS is a colour problemwhich affects the ultimate discharge of treated effluent (l3attersby, 1988).Callander and Barford (1983) and Sarner et al. (1988) concluded that adding ironsalts in order to remove sulphide from solution should be ruled out as a viable long termsolution. These recommendations as well as preliminary laboratory experience with ironsulphide precipitates in anaerobic reactors force the consideration of another method ofsuiphide removal from the reactor liquor.Gas StrippingThe driving force for hydrogen suiphide transport from the liquid to the gaseousphase is the difference in the partial pressure of hydrogen suiphide in equilibrium with theliquid and the actual partial pressure ofH2S in the gas (Winfrey and Zeikus, 1977, Sarner161et a!., 1988). Hydrogen suiphide free gas recirculation through the reactor liquor utilizesthe gas liquid equilibrium which is governed by Henry’s law:jH2S]aq = ajH2SJg (2)The absorption coefficient, a, is 1.83 at 35 °C (Lawrence et aL, 1966) and 1.59 at55 °C (Wilhelm et al., 1977). When the partial pressure of hydrogen sulphide is lowered inthe gas phase by gas scrubbing, the equilibrium shifts to the right, drawing H2S out ofsolution. Hydrogen sulphide is alone among the suiphide forms which partitions betweenthe liquid and gaseous phases. Sarner et at. (1988) neatly summarized the transport ofsulphide in an anaerobic reactor as:H2S gifH2S aq HS +When hydrogen sulphide is transferred from the liquid to the gas, the dissociationequilibrium shifts to the left, from the ionic to the free suiphide form. Thus, continuouslyrecirculating sulphide-free methane and carbon dioxide gas through the reactor liquorforces the dissolved sulphide into the gaseous phase where it can be removed byscrubbing. It is expected that recirculating methane and carbon dioxide gases through thereactor liquor should have little effect on their equilibrium concentration in solution.A number of alternatives are available for H2S removal from the biogas. For thisapplication, the literature cites ferric chloride (Oleszkiewicz and Hilton, 1987) and zincacetate (Isa et aL, 1986, Oleszkiewicz and Hilton, 1987) along with various techniques toregenerate the scrubbing solutions and to recover the sulphur (Maree and Strydom, 1983,Jensen et aL, 1988, Buisman eta!., 1990).Hydrogen sulphide removal systems of this type have been well researched by Isaet at. (1986) and Hilton and Oleszkiewicz (1987) for example. These investigators foundthe performance of the gas stripped reactors to surpass that of the unstripped reactors andto equal that of the control reactors without sulphate added. Sulphate reduction to162suiphide was enhanced in the stripped reactors, pH stability was improved due to thenatural alkalinity generated by suiphidogenic reactions and inhibition was rapidly andcompletely reversible within a short time (Hilton and Oleszkiewicz, 1987).Removing sulphide as hydrogen sulphide gas rather than forming a sulphideprecipitate in the reactor has a number of advantages:a) Both the SRB and the MPB contribute to the TOC treatment efficiency, unconstrainedby elevated sulphide levels in the stripped reactor.b) Since the H2S removal system is outside of the reactor system, it is not constrained bythe limits of the digester environment, especially pH. Also, a dedicated stripper can beoperated independent of biomass retention considerations.c) A sludge precipitate is not produced in the reactor. Thus, the active reactor volume isnot diminished by the accumulation of precipitate.d) The scrubbing solution may be regenerated, minimizing on-going chemical costs.Sulphur recovery is possible by physical-chemical (Sarner, 1986) or by biological(Buisman et al., 1989) processes. This could be of economic importance to some mills.e) There are no added corrosion problems (as there are for example with ferric chloride)although the sulphide itself does impose constraints on materials selection.f) The liquid effluent is not additionally discoloured.g) Sulphide is removed from the liquid effluent and does not add to the oxygen demand.Disadvantages of gas stripping include:a) The energy costs for gas recirculation are high (Jensen et aL, 1988).b) The installation and operation of an auxiliary hydrogen sulphide stripping and scrubbingsystem and the disposal or regeneration of the spent scrubbing solution is an addedcomplication and expenditure.c) The increased gas flow could buoy up suspended biomass, causing it to be wasted fromthe reactor and thus decreasing from system capacity and stability or adding to the clarifierload. Consequently, a dedicated liquid/solid separation unit would be required.163d) Increased reactor mixing disrupts the quasi plug flow characteristics and thus thesubstrate concentration profile in the reactor. This short circuiting leads to higher effluentconcentrations exiting from the reactor.e) An alkaline pH promotes the SRB activity but any reactor pH above 6.4 detracts fromthe stripping efficiency since stripping is effective only when the sulphide is present ashydrogen sulphide.9.3 Results: Effect of Hydrogen Suiphide Stripping at 35 °CTwo different effluent batches were used in these gas stripping experiments. Thefirst batch was used to perform a control run and a gas stripping run. The control run(“run OH) employed drawing gas from the head space through a sulphide saturated waterscrubber and then recycling the gas through the liquor of the methane phase reactor. Thisperformance was compared with gas scrubbing through a 100 g/l ferric chloride solution(replaced daily) with suiphide free gas recycled through the methane phase reactors (“run1“) to strip the hydrogen suiphide out of solution. Subsequent experiments used 200 g/lferric chloride in the scrubbers in a successful attempt to improve the removal of hydrogensuiphide. The second effluent batch was low in sulphate concentration. Run 2 wasperformed with gas scrubbed through ferric chloride using a low sulphate feed. Asupplementary run (“run 3”) was performed with influent sulphate levels boosted byadding sodium sulphate to levels comparable to the other experiments.a) Total Dissolved SulphideScrubbing hydrogen sulphide by passing the head space gases through a ferricchloride solution and then stripping the methane phase reactor liquor by means ofrecycling the hydrogen suiphide free gas through the MPRs was effective in decreasing thetotal dissolved sulphide concentrations. The total dissolved suiphide concentrations of the164stripped methane phase reactors were significantly decreased compared to the control run(see Table 40). More than half of the total dissolved sulphide was removed in the ferricchloride scrubbing experiment, batch 1, compared to the control (gas recycle through awater scrubber) from 186 to 266 mg/I sulphide without scrubbing compared to 69 to 107mg/i sulphide with stripping. The increased infiuentsulphate levels in batch 2 comparedtGthe unsupplemented batch 2 effluent led to higher dissolved sulphide levels in the MPR.This indicates that the bio-reactor/stripping systems were quite limited in their capacity.It is expected that the sulphide removal efficiencies recorded here could beincreased by designing the reactors to be more amenable to gas stripping or by installing astripping vessel between the acid phase reactor and the methane phase reactor. The key toremoving suiphide from solution would be to operate the stripper at a low as possible pHin order to maximize the fraction of un-ionized sulphide which is partitioned between thegaseous and aqueous phases. Stripping at acidic pH levels would be very compatible withthe operation of the acid phase reactors. Earlier experiments however, indicated theefficacy of operating the acid phase reactors under slightly alkaline conditions. Other suchstripping improvements would center around increasing the gas contact with the liquor bymeans of fine bubble diffusers in the reactor and by increasing the flow rate of the gasrecycle. Of course, the retention of active biomass in the presence of high gas flow rateswould require dedicated liquid solids separation equipment. Alternatively, a dedicatedstripper system could offer the advantages of independence of pH level and circumvent thedifficulties of biomass retention and control.The total dissolved suiphide concentrations increased with increasing HRT from0.8 to 1.2 days for all of the scrubbing experiments. There was no significant difference inthe mean total dissolved suiphide concentrations between the 1.2 and 1.8 day total HRTruns.Differences between the dissolved suiphide levels of any of the MPRs were notexpected since the rate of suiphide removal in the chemical scrubber could be expected to165be far greater than the rate of suiphide formation by the SRB. That there were differencesin suiphide levels between the reactors based on HRT points to the limitations withstripping these MPRs. The three methane phase reactors were of identical height but ofvarying widths to accommodate the different reactor volumes. Thus, with the greatestaspect ratio, the total 0.6 day HRT (smallest) reactor system was best suited to stripping.Also, since the gas recycle rate for all three reactors was the same, the 0.6 day HRTreactor had the greatest recycle ratio (gas recycle rate:gas production rate or gas recyclerate:liquid feed rate). These two design differences favour the stripping efficiency of thesmallest HRT reactor over the larger reactor systems. Rather than employing the samestripping rate as was done for all three reactor sets, a consistent specific gas recycle rate (1gas/minll reactor) would have been a better design decision.Table 40: Effect of Hydrogen Suiphide Stripping on Total Dissolved SuiphideConcentration at 35 °CMean Total Dissolved Suiphide Concentration ± SD 1mg/liGas Feed Acid/Methane Phase Reactor EERT [days]Stripping S2 APR MPR APR MPR APR MPR±SDExperiment [mg/I] 0.2 0.4 0.4 0.8 0.6 1.2water scrubber(control) 3.1±0.3 132± 15 186±6 211± 14 264±9 170± 13 266± 15ferric chloridescrubberbatchi 3.1±0.2 90± 12 69± 12 102±9 108± 14 126± 17 107±7ferric chloridescrubberbatch2 3.3±0.2 44±5 24±4 53±5 27±3 54±4 32±3ferric chloridescrubberbatch2 3.2±0.1 69± 11 58±9 103± 12 92±23 99± 10 93±9+ sulphate166H2S gas was assayed in these gas scrubbing experiments. The control rundemonstrated hydrogen sulphide concentrations in the head space of between 2.0 and2.5% by volume. The ferric chloride scrubbing experiments effectively removed hydrogensulphide from the gas phase. No hydrogen sulphide gas was detected in these experiments.The second effluent--batch, when -boosted with sodium -sulphate, -produced. resultswhich were not significantly different from the stripped experiments with the previouseffluent batch. Screening the effects of stripping, HRT and influent sulphate levels, linearregression analysis revealed the influent sulphate concentration as a significant variableaffecting the sulphide concentration in the methane phase reactor.b) SulphateTable 41 summarizes the effects on sulphate of stripping the methane phasereactors with sulphide-free, recirculated biogas at 35 °C. The sulphate reduction efficiencyfor the methane phase reactors is an overall efficiency and it includes the sulphate removedin the acid phase reactors.Sulphate reduction appeared to be unaffected by gas stripping, although the largevariability in the sulphate concentrations, for run 1 in particular preclude confidentpredictions of the effects of scrubbing for these experiments. That the SRB were notstrongly affected by even the high levels of total dissolved sulphide confirms their robustnature as has been reported in the literature. Isa et aL (1986) and Maillacheruvu et aL(1993) also observed sulphate reduction to be independent of suiphide levels. Addedsulphate to batch 2 did not significantly affect the sulphate reduction efficiency.The sulphate reduction efficiency was the lowest at the smallest HRT but therewas no significant difference between the 1.2 and 1.8 day total HRT systems.Consistent with the above observations, the effects of stripping, influent sulphatelevels and HRT on the sulphate reduction efficiency in the methane phase reactors are alsosummarized in the linear regression equation:167%SO4rMpR 54+ + 0.12*SMpR - O.O34*SO4ApRr=0.89 p=O.OO1In order to assess the mixing effect of gas recycle through the acid phase reactoron sulphate reduction, a comparison can be made to the acid phase reactor pHexperiments (see Tab1e 4)No significant difference isevident, indicating either.mixing of the MPB did not greatly affect the extent of sulphate reduction or that thedifferences in the wastewater composition make the validity of such a comparisondoubtful.Table 41: Effect of Hydrogen Suiphide Stripping on Sulphate Removal at 35 °CMean Sulphate Reduction Efficiency j%] ± CVGas Feed Acid/Methane Phase Reactor HRT [days]Stripping SO42- APR MPR APR MPR APR MPR±SDExperiment [mg/i] 0.2 0.4 0.4 0.8 0.6 1.2water scrubber(control) 1400±3 7±8 30±10 41± 11 87± 18 28±8 85±6femc chloridescrubberbatchi 1410±3 6±5 20±7 29± 15 64±23 23±8 80±28ferric chloridescrubberbatch2 570±2 69±5 71± 15 67± 11 70± 15 64±5 70± 14ferric chloridescrubberbatch2 1170±9 38± 11 56± 10 43± 12 72± 14 47± 14 83± 12+ sulphatec) Suiphite and ThiosuiphateStripping suiphide from the MPR substantially changed the suiphite concentrationsin the first wastewater batch (see Table 42). Whereas the control run demonstratedincreased sulphite concentrations in the methane phase reactors compared to the acidphase reactors, a decrease of dissolved sulphide caused these levels to be substantially168diminished. This effect was not consistently observed for the second wastewater batch forthe stripping experiments.Linear regression analysis showed that the sulphite concentrations measured in themethane phase reactor were independent of HRT and influent sulphate concentration butwere dependent upon the stripping effectiveness, as measured by the dissolved sulphideconcentration, and upon the suiphite concentration exiting the acid phase reactor. Theequation is as follows:SO3MpR = -17 + 0.13*S2MpR + 0.96*SO3AJJRr=0.76 p=0.0I6A number of investigators have observed that suiphite is not prone to reduction byanaerobic wastewater treatment microorganisms. From the results produced in this work,it appears that elevated suiphide concentrations may have a role in blocking the reductionof sulphite. In all of these stripping experiments, the measured suiphite concentrationswere less than what other investigators have found to be problematic in anaerobicwastewater treatment.The thiosulphate trends with gas stripping are presented in Table 43. Due to thescarcity of reports, whether or not these thiosuiphate levels are typical or of concern isunknown.High thiosulphate levels coincided with elevated suiphide concentrations. Thecontrol run found the thiosulphate present in the effluent to be roughly tripled in thereactors. The thiosulphate concentrations in the methane phase reactors were onlymarginally higher than in the acid phase reactors. At these HRTs, there was no simplerelationship between the reaction time and the thiosuiphate level. Stripping led tosomewhat lower thiosuiphate concentrations in the methane phase reactors with thedecrease in thiosuiphate concentration positively correlated with longer HRTs.169Table 42: Effect of Hydrogen Suiphide Stripping on Suiphite Concentrationat 35 °CMean Sulphite Concentration ± SD [mg/I]Gas Feed Acid/Methane Phase Reactor BRT [days]Stripping SO32- APR MPR APR MPR APR MPR±SDExperiment [mg/i] 0.2 0.4 0.4 0.8 0.6 1.2water scrubber(control) 27±2 22±7 43±7 16±3 45±4 20±8 22±8ferric chloridescrubber batch 1 31±7 26±4 4±2 20±5 6±2 15±2 7±2ferric chloridescrubber batch2 19±2 25±5 18±2 19±3 12±2 22±8 13±4ferric chloridescrubber batch 2 11±3 12±4 0±0 8±3 0±0 9±2 5±1+ sulphateTable 43: Effect of Hydrogen Suiphide Stripping on Thiosulphate Concentrationat 35 °CMean Thiosuiphate Concentration ± SD [mg/I]Gas Feed Acid/Methane Phase Reactor HRT [days]Stripping S203 APR MPR APR MPR APR MPR±SDExperiment [mg/I] 0.2 0.4 0.4 0.8 0.6 1.2water scrubber(control) 48±18 151±14 165±18 130±7 156±14 162±13 174±10ferric chloridescrubberbatchl 41±11 129±23 104±14 161±17 86±23 150±6 41±8ferric chloridescrubberbatch2 8±2 41±7 8±3 23±8 10±3 32±2 18±6ferric chloridescrubberbatch2 4±1 57±18 62±6 73±12 74±16 77±13 76±11+ sulphate170The second wastewater batch was characterized by much lower thiosuiphatelevels. Stripping decreased thiosuiphate levels compared to those in the acid phasereactors. Increased influent sulphate levels which resulted in higher suiphide levels alsoresulted in raised thiosuiphate concentrations.All of these results are summarized in the linear regression equation. The effect ofsuiphide stripping, influent sulphate and thiosuiphate concentrations, and HRT werescreened in a stepwise regression procedure, resulting in the following relationship:S2O3MpR = -7.7 + 0. JO*SO4AFR + 0.63*S2R-0.60 *S2Q3APRr=0.97 p=O.000d) Wastewater Treatment EfficiencyHydrogen suiphide gas scrubbing resulted in the highest TOC removals comparedto the other sulphur management experiments (see Table 44). Since lower suiphideconcentrations did not enhance sulphate reduction (Table 41), the improved treatmentefficiency appears to be due to a decreased suiphide inhibition of the non-SRB acetogensand the MPB.In general, longer HRTs led to enhanced treatment efficiencies. In the firstwastewater batch, lowered dissolved sulphide concentrations led to substantial gains in theTOC removal efficiency, from a maximum of 24% removal in the control runs to amaximum of 57% in the stripped reactors.In the second wastewater batch, elevated sulphate levels in the influent resulted inan improved TOC removal efficiency. This appears to be due to the increasedconsumption of organic compounds in the course of sulphate reduction. It would appearfrom these experiments that while sulphate produces suiphide which impacts negatively onTOC removal, the reduction of sulphate without the accumulation of sulphide is of benefitin removing organic compounds from the waste stream. These results support the171observations of Hilton and Oleszkiewicz (1987) who found that a syntrophic generation ofCH4 and H2S were essential to achieve high effluent reductions.Table 44: Effect of Hydrogen Sulphide Stripping on TOC Removal at 35 °CMean TOC Removal Efficiency j%] ± 5 % CVGas Feed Acid/Methane Phase Reactor HRT [days]Stripping TOC APR MPR APR MPR APR MPRExperiment [mg/I] 0.2 0.4 0.4 0.8 0.6 1.2water scrubber(control) 2990 15 22 22 24 17 24ferric chloridescrubberbatchi 3245 28 37 29 43 34 57ferric chloridescrubberbatch2 3475 31 32 41 47 30 47ferric chloridescrubber batch 2 3540 29 43 38 51 41 56+ sulphateThe effects of HRT, sulphide stripping and influent sulphate concentration on theTOC removal efficiency were assessed using stepwise linear regression. All three variableswere revealed to be statistically significant as follows:%TOCrMpR =29+29 *HRTMpR + 0.01 *SO4ApR - 0.14*S2MpRr=0.89 p=O.OO4e) Gas ProductionFor the scrubbing experiments with the first batch of effluent, ferric chloride gasscrubbing promoted an increase in gas production in one reactor whereas the effects in theother two reactors were not statistically significant (see Table 45). This suggests that theelevated dissolved suiphide levels in the control run inhibited the gas yielding reactions ofthe microorganisms. A decrease in these sulphide concentrations was brought about bygas scrubbing and stripping alleviated this inhibition to some extent. The second effluent172batch considered the effects of changing sulphate concentration. Increased sulphateconcentrations increased the total gas production rate but the methane content decreasedfrom 52 to 43%.Table 45: Effect of Hydrogen Suiphide Stripping on Gas Production Rate at 35 °CGas Mean Gas Production Rate ± SD [I/day]Stripping Acid+Methane Phase Reactor HRT [daysiExperiment 0.6 1.2 1.8water scrubber(control) 0.30±.04 0.64±.02 1. 14±.06ferric chloridescrubber batch 1 0.25±.04 1.46±.25 1.20±.22ferric chloridescrubber batch 2 0.20±.05 1.07±. 18 1.85±. 11ferric chloridescrubberbatch2 0.31±,06 1.39±.14 2.30±.23+ sulphate1) Volatile Fatty AcidsVolatile fatty acids are the intermediate organic substrates for the SRB, the nonSRB acid forming bacteria, and the methanogens. Their mean concentrations for the gasscrubbing experiments are presented in Figure 8. An explanation for the trends which wereobserved for the gas scrubbing and stripping experiments conducted at 35 °C follows.Formic Acid: Formate was reduced to near zero levels at all HRTs for both the acidphase reactors and methane phase reactors for each of the experiments. Utilized by theMPB, the disappearance of formate merely confirms the occurrence of somemethanogenic activity.Acetic Acid: Acetate concentrations in both the acid phase reactors and methane phasereactors were always greater than in the feedstock. Stripping of hydrogen sulphide appears173to have had little effect on the acetate concentration in the methane phase reactors at anyof the HRTs tested. This suggests that the SRB and the non-SRB acetogens were notstrongly inhibited by sulphide at the levels examined and that inhibition of the acetate-consuming methanogens was only weakly alleviated by decreasing the total dissolvedsulphide concentrations from approximately 260 to 100 mg/l sulphide. Inhibition of theacetoclastic MPB may have been initiated by even lower levels of sulphide or bycompounds other than sulphide. The results of run 3 which repeated the previous run 2but at 1170 rather than 570 mg/i sulphate demonstrate that there appears to be nosignificant effect of sulphate concentration on net acetic acid production or consumption.This confinns the literature findings of the robust nature of both the SRB and the nonSRB acid producing bacteria.Stepwise linear regression analysis bears out these qualitative observations.Acetate concentrations obtained from the methane phase reactor were found to be lowerat increased HRT, slightly increased with higher dissolved sulphide levels, and stronglyrelated to the acetate levels of the acid phase reactor. The changing sulphateconcentrations had no statistically significant effect on acetate levels in the methane phasereactor. The regression equation was calculated to be as follows:AcetateMpR — 550 37Q *HRTMpR + J.5*S2MpR + 0.9*Acetate APRr=0.995 p=O.000Propionic Acid: Propionic acid is used by both the SRB as well as the non-SRBacetogens. In the absence of scrubbing, a net production of propionic acid was observed.Stripping suiphide from the reactor liquor led to an increase in the concentration ofpropionic acid. Since scrubbing resulted in lower dissolved suiphide concentrations but notin enhanced sulphate reductions, it would appear that the elevated suiphide levels in thecontrol run retarded the generation of propionate by non-SRB acetogens.1742502001500-__.0 .e1: zoo.50o----*----4 o,0 1 2 3 0 1 2 3Gas Scrubbing Experiment Gas Scrubbing Experiment250 100-E200I1 75.150 __—:___Gas Scrubbing Experiment Gas Scrubbing Experiment500400 —— Feed0 0.2dHRTAPR1d 300 0.4 d HRT APR 2o 200 0----- 0.6 d HRT APR 3—.—0.4dHRTMPR1100•—A—0.8 d HRT MPR 2I--:• 1.2 d HRT MPR 30 1 2 3Gas Scrubbing ExperimentFigure 8: Effect of Gas Stripping at 35 °C on Volatile Fatty Acids175Runs 2 and 3 considered the effects of varying the sulphate concentrations on thestripped reactor system. Low influent sulphate levels resulted in large increases in thepropionic acid concentration compared to the feed. With small sulphate levels supportinglittle SRB activity, propionic acid was sparingly consumed by the SRB and excesspropionate was generated. Doubled sulphate concentrations resulted in a large decrease inthe propionic acid concentrations for each of the reactors. This appears to havedemonstrated the utilization of propionate by the SRB.Butyric Acid: Feed n-butyric acid concentrations were significantly decreased in the acidphase reactors. In general, the n-butyric acid concentrations in the methane phase reactorswere not significantly different from the acid phase reactor concentrations. Neitherscrubbing nor HRT appear to have had an effect on the n-butyric acid concentration in themethane phase reactors. Stepwise linear regression analysis showed that n-butyric acidlevels of the methane phase reactors were not linearly related to HRT or suiphideconcentrations, but rather were dependent upon the in-coming butyrate concentrations andon the sulphate levels of the influent.Added sulphate appears to have had an effect on the n-butyric acid concentrations.N-butyric acid levels decreased to near zero levels with sulphate addition. This may havebeen due to the increased activity of the SRB which utilize n-butyrate as a carbon sourceto reduce sulphate to sulphide.Lactic Acid: Feed lactate concentrations were decreased significantly in the reactors forall of the scrubbing experiments. No significant difference was measured at any HRTbetween the acid phase reactors and the methane phase reactors. The effect of sulphateaddition on lactate concentration could not be determined since both runs resulted in nearzero lactate concentrations.176g) Reactor pHTable 46: Effect of Hydrogen Suiphide Stripping on Reactor pH at 35 °CMean Reactor pH ± SDGas Feed Acid/Methane Phase Reactor HRT jdaysjStripping pH APR MPR APR MPR APR MPRExperiment ± SD 0.2 0.4 0.4 0.8 0.6 1.2water scrubber(control) 7.O±.2 7.l±.l 7.l±.2 7.l±.l 7.2±.l 7.l±.l 7.2±.lferric chloridescrubberbatch 1 6.8±.l 7.1±.1 7.2±.1 7.O±.1 7.3±.2 7.O±.1 7.3±.1femc chloridescrubber batch2 6.8±.1 7.O±.l 6.7*.2 7.O±.1 7.3±,l 7.O±,l 7.3±.1ferric chloridescrubber batch2 6.75±.14 7,O±.1 7.O±.l 7.O±.l 7.6±.2 7.O±.l 7.4±.1+ sulphateTable 46 summarizes the effect of the stripping experiments on the pH of thereactors. Longer HRTs and the reduction of elevated concentrations of sulphate in thestripped methane phase reactors resulted in a significant MPR pH increase over that of theAPRs. These trends have been reported in the literature (Middleton and Lawrence, 1977,Oleszkiewicz and Hilton, 1977).h) Carbonate ConsumptionWhile stripping was performed in the methane phase reactors, pH control was onlyeffected in the acid phase reactors. Consequently, within each wastewater batch, there wasno significant difference in carbonate consumption for the stripped versus the unstrippedreactors. As for the other experiments, an increase in APR HRT from 0.2 to 0.4 days ledto an increased consumption of sodium carbonate but little difference was noted between1770.4 and 0.6 days. The varying sulphate levels also did not markedly affect carbonateconsumption at the levels tested. These results are presented in Table 47.Table 47: Effect of Hydrogen Suiphide Stripping on Carbonate Consumptionat 35 °CGas Mean Sodium Carbonate Consumption (gil feed]Stripping Acid Phase Reactor HRT (days]Experiment 0.2 0.4 0.6water scrubber(control) 0.61 1.37 1.34ferric chloridescrubberbatchi 0.60 1.19 1.14femc chloridescrubber batch 2 1.07 1.84 1.98ferric chloridescrubberbatch2 0.95 1.91 1.82+ sulphate9.4 Results: Effect of Hydrogen Suiphide Stripping at 55 °Ca) Total Dissolved SuiphideThe total dissolved sulphide concentrations of the stripped methane phase reactorswere significantly decreased compared to the unscrubbed control runs (see Table 48). Themethane phase reactor sulphide levels in the range of 125 mg/i were decreased toapproximately 50 mg/I after scrubbing. HRT had no significant influence on the totaldissolved sulphide levels after scrubbing. The 0.6/1.2 day acid/methane phase reactorfunctioned poorly for this set of experiments due to an uncontrolled sodium carbonateaddition at the end of a prior experiment.Hydrogen sulphide gas was removed in the 200 gIl ferric chloride scrubbers toundetectable levels. The mean hydrogen suiphide gas concentrations in the hydrogen178suiphide saturated water scrubber (mixed but unscrubbed) control were between 1.6 to2.2% by volume.Table 48: Effect of Hydrogen Suiphide Stripping on Total Dissolved SuiphideConcentration at 55 °CMean Total Dissolved Suiphide Concentration ± SD [mg/liGas Feed Acid/Methane Phase Reactor HRT [daysjStripping SO42- APR MPR APR MPR APR MPR±SDExperiment [mg/lI 0.2 0.4 0.4 0.8 0.6 1.2water scrubbercontrol 1580±30 83±14 125±7 92±1 124±13 22±5 71±7ferric chloridescrubber 1540±76 22±3 52±8 95±15 52±8 25±4 48±10b) SulphateStripping total dissolved sulphide from the reactor liquor had no significant effecton sulphate reduction at 55 °C (see Table 49). Changing the HRT from 0.4 to 1.2 daysalso did not affect the extent of sulphate reduction. Such an apparent lack of inhibition ofthe SRB by sulphide was also observed for the mesophilic experiments. The sulphatereduction efficiency was low for this set of experiments, ranging from only 32 to 42%.The possible explanations for the low sulphate reduction which was observed throughoutthe thermophilic experiments have been presented in earlier sections. This suggests thatthe SRB were not any more inhibited at 125 mg/i sulphide than they were at 50 mg/isuiphide.c) Suiphite and ThiosulphateTable 50 summarizes the measured suiphite concentrations for the thermophilicstripping experiments. Sulphite concentrations were unchanged by stripping the methane179phase reactor at 55 °C and were not significantly different from the feed. For the twolower HRT reactor systems, the HRT did not affect sulphite concentrations at the levelstested. The sulphite concentrations were small enough so as to not significantly impact onthe wastewater treatment microorganisms.Table 49: Effect of Hydrogen Suiphide Stripping on Sulphate Reduction at 55 °CMean Sulphate Reduction Efficiency [%] ± CVGas Feed Acid/Methane Phase Reactor HRT [days]Stripping SO42- APR MPR APR MPR APR MPR±SDExperiment [mg/I] 0.2 0.4 0.4 0.8 0.6 1.2water scrubbercontrol 1580±2 23±5 36±2 24±4 38±4 5±1 22±4ferric chloridescrubber 1540±5 32±4 34±7 34±8 42±7 10±5 27±8Similar to the previous sulphur management experiments, thiosuiphateconcentrations in the reactors were greater than in the feedstock, again suggesting thatthiosulphate is an intermediate compound in the reduction of sulphate. No statisticallysignificant relationship between HRT or scrubbing and thiosuiphate concentration couldbe established. The previously upset 0.6/1.2 day HRT reactor system demonstrated lowthiosulphate concentrations. Thiosuiphate measurements for these experiments aresummarized in Table 51.180Table 50: Effect of Hydrogen Suiphide Stripping on Suiphite Concentrationat 55 °CMean Suiphite Concentration ± SD [mg/i]Gas Feed Acid/Methane Phase Reactor HRT [days]Stripping So32- APR MPR APR MPR APR MPR±SDExperiment [mg/I] 0.2 0.4 0.4 0.8 0.6 1.2water scrubbercontrol 23±5 15±3 20±1 18±2 20±1 10±3 15±2ferric chloridescrubber 15±4 13±3 19±4 18±4 17±3 14±5 18±4Table 51: Effect of Hydrogen Suiphide Stripping on Thiosuiphate Concentrationat 55 °CMean Thiosuiphate Concentration ± SD [mg/I]Gas Feed Acid/Methane Phase Reactor HRT [days]Stripping S2032 APR MPR APR MPR APR MPR±SDExperiment 1mg/il 0.2 0.4 0.4 0.8 0.6 1.2water scrubbercontrol 3±1 36±7 45±15 108±13 86±12 3±1 16±7ferric chloridescrubber 0±0 26±2 54±4 93±20 23±7 5±3 13±5d) Wastewater Treatment EfficiencyScrubbing and stripping of hydrogen suiphide from the methane phase reactorsresulted in enhanced wastewater treatment efficiencies, from 27 to 32% (water scrubbercontrol) to 36 to 39% (ferric chloride scrubber). Table 52 summarizes these results. TheTOC removal efficiency for the methane phase reactors is an overall efficiency and itincludes the TOC removed in the acid phase reactors. The TOC removal was only181marginally improved by increasing the nominal total HRT from 0.6 to 1.8 days in eitherthe control or ferric chloride scrubbing experiments.Regression analysis was consistent with these observations. Stepwise linearregression revealed both the lowered sulphide levels by stripping and HRT to exert asignificant effect on the extent of TOC removal. The regression is as follows:%TOCrMpR 53- O.OO8*HRTp - O.]3*S2MPRr=O.99 p=O.OO.2Table 52: Effect of Hydrogen Suiphide Stripping on TOC Removal at 55 °CMean TOC Removal Efficiency j%1 ± 5 % CVGas Feed Acid/Methane Phase Reactor HRT LdaysiStripping TOC APR MPR APR MPR APR MPRExperiment [mg/lI 0.2 0.4 0.4 0.8 0.6 1.2water scrubbercontrol 2940 27 27 21 28 7 32ferric chloridescrubber 2940 27 37 34 39 26 36As with the mesophilic experiments, this system of hydrogen suiphide gasscrubbing and stripping resulted in the highest TOC removals compared to the othersulphur management experiments.Since sulphide removal did not markedly enhance sulphate reduction (Table 49),the improved treatment efficiency cannot be attributed to changes with the SRB. Theimproved carbon removal efficiencies with sulphide scrubbing point to some degree ofsuiphide inhibition of the MPB at 125 mg/l sulphide, the concentration measured in theunstripped control run. Whether or not increased scrubbing efficiency, leading to lowertotal dissolved sulphide concentrations than achieved here, would result in enhanced TOC182removals remains an open question. Decreasing the total dissolved sulphide concentrationsto below the 50 mg/i level attained in these experiments may be of dubious benefit sincethere are few reports of suiphide inhibition of the MPB at this level.e) Gas ProductionThe mean gas production rates are shown in Table 53. Lowering the totaldissolved hydrogen suiphide levels in the methane phase reactors increased the rate oftotal gas production. Gas production was also affected by HRT, with longer NRTsgenerally resulting in higher gas yields (r 0.75).Table 53: Effect of Hydrogen Suiphide Stripping on Gas Production Rate at 55°CGas Mean Gas Production Rate ± SD [I/day]Stripping Acid+Methane Phase Reactor HRT [days]Experiment 0.6 1.2 1.8water scrubbercontrol 0.10±.05 0.07±.06 0.30±07ferric chloridescrubber 0.15±. 10 0.17±.07 0.52±.031) Volatile Fatty AcidsThe effects of stripping the MPRs of dissolved suiphide on the individual VFAlevels in the reactors are briefly reviewed below.Formic Acid: Formate concentrations were reduced to near zero in the normallyfunctioning reactors at all HRTs and for both the ferric chloride scrubbing experimentsand the control.The reactor which had been upset at the end of the previous experimentdemonstrated a reduced capacity to utilize formate than did the other two stable reactorsystems.183Acetic Acid: Already high influent acetate levels were increased further in the APRs andalso in the MPRs. As the HRT decreased, the acetate concentration increased, indicativeof a rapid rate of acetogenesis by either the SRB or the non-SRB acid forming bacteriaand a slow rate of acetate utilization. Removal of the total dissolved sulphide viascrubbing slightly increased the acetate concentrations. Since the -MPB were unable toefficiently use acetate as demonstrated by the low rates of methane production, theincreased acetate concentrations may have been due to an increased metabolism of theacetate producing SRB or of the non-SRB acetogens at lowered dissolved sulphide levels.Propionic Acid: Propionic acid concentrations were greater in the reactors than in thefeed. In the unscrubbed reactors, propionic acid levels, which were generated in the APRs,were increased in the MPRs. Scrubbing lowered the total dissolved sulphide concentrationin the MPRs and led to lower propionate concentrations. This suggests inhibition atelevated suiphide levels of the microorganisms which consume propionic acid. Suchmicrobes include the SRB as well as the non-SRB acetogens.Butvric Acid: Butyric acid concentrations were low in the feedstock and were notsignificantly different in the normally functioning reactors. Regression analysis showedthat the butyric acid levels were not linearly related to either HRT or the lowered suiphidelevels in the methane phase reactor. The previously upset reactor demonstratedconsistently high n-butyric acid levels in the APRs but these concentrations were reducedto near zero in the MPR.Lactic Acid: Lactic acid was consumed in the reactors. Iniluent lactic acid concentrationswere considerably lower in the scrubbed reactors compared to the control run. Thisconfounds the ability to describe the effects of scrubbing on lactic acid levels. Stepwiselinear regression analysis revealed the effluent lactic acid concentrations to simply bedependent on the lactate concentrations in the acid phase reactors and not linearly relatedto either HRT or suiphide removal by stripping.1842000I::: E::::;100o 50050-0control scrubbing Control ScrubbingGas Scrubbing Experiment Gas Scrubbing Experiment250I::: .::::::z1100........... Icontrol scrubbing control scrubbingGas Scrubbing Experiment- Gas Scrubbing Experiment500400 —— Feed0.2dHRTAPR1O.4dHRTAPR2200•-•- 0.6 d HRT R 3100—•— 04 d HRT MPR 1—A-—ogdmTMpR20 O 1.2dHRTMPR3control scrubbingGas Scrubbing ExperimentFigure 9: Effect of Gas Stripping at 55 °C on Volatile Fatty AcidsL 85g) Reactor pHThere was no significant difference in the MPR pH levels with or withoutscrubbing (see Table 54). The MPR pH levels fell to less than neutral both with andwithout removal of dissolved sulphide, but were still within an acceptable range for theMPB to function. The decreased pH may be due to greater VFA production in the absenceof large sulphide concentrations or directly arising from the sulphide concentrations orarising from some carry over of ferric chloride from the scrubber to the reactor liquor.Disregarding the previously upset 0.6/1.2 day HRT reactor system, an increase inHRT led to an increase in reactor pH, likely due to the increased sulphide levels whichwere generated. A similar trend was observed at the 35 °C stripping experiments. Noeffect of stripping is evident from these experiments.Table 54: Effect of Hydrogen Suiphide Stripping on Reactor pH at 55 °CMean Reactor pH ± SDGas Feed Acid/Methane Phase Reactor HRT [days]Stripping pH APR MPR APR MPR APR MPRExperiment ± SD 0.2 0.4 0.4 0.8 0.6 1.2water scrubbercontrol 5.6±.4 7.2±.l 6.8±.l 7.O±.1 7.1±.1 7.1±.l 6.5±.2ferric chloridescrubber 5.5±.1 7.2±.1 6.7±.1 7.1±.1 7.2±.2 7.1±.1 6.8±.2h) Sodium Carbonate ConsumptionThe mean sodium carbonate consumption was not significantly different for theferric chloride scrubbed runs as shown in Table 55, since carbonate was added to the acidphase reactors whereas scrubbing was performed on the methane phase reactors.186Table 55: Effect of Hydrogen Suiphide Stripping on Carbonate Consumptionat 55 °CGas Mean Sodium Carbonate Consumption Fgfl feed]Scrubbing Acid Phase Reactor HRT IdaysiExperiment 0.2 0.4 0.6water scmbbercontrol 0.95 1.23 0.51fernc chloridescrubber 0.98 1.42 0.779.4 Conclusions: Effect of Sulphide StrippingAt 35 °C, recycling hydrogen sulphide-free gas through the methane phase reactorresulted in decreased suiphide concentrations, roughly half of the control experiment. Thehigh suiphide levels in spite of stripping point to the limitations of stripping inside of thebio-reactor. The SRB were not significantly affected by the decreased sulphide levels. TheTOC removal was greatly improved in the stripped methane phase reactors, from 22 to24% in the control to 37 to 57% in the stripped reactors.Elevated sulphate levels in the feed led to increased sulphide concentrations in themethane phase reactor but also to a small increase in the TOC removal efficiencies,suggesting that the SRB contributed to the overall wastewater treatment efficiency.At 55 °C, stripping the methane phase reactor decreased the sulphide levels, hadno effect on the extent of sulphate reduction, and resulted in marginally improved TOCremoval efficiencies.Similar to the mesophilic work, thermophilic operation resulted in high acetateconcentrations and low methane production rates. It is apparent that the lowconcentrations of both sulphide and sulphate were not sufficient to alleviate inhibition of187the methane forming bacteria. Other compounds in the effluent, such as the various woodextractives or DTPA, may have accounted for this inhibition.188Chapter 10Effect of Temperature, Time and Phase Separation10.1 OverviewSection 2 reorganizes the results to describe the effects of temperature on sulphatereduction, suiphide concentration, TOC removal, gas production and VFAievel& -Section 3 describes the startup of the reactor system after a period of dormancy.Section 4 describes the effect of HRT on the wastewater treatment parameters.Section 5 considers the effect of phase separation on reactor performance.Section 6 compares similar reactor settings throughout the experimental programto estimate the reproducibility of the results.10.2 Effect of Temperature (35 °C versus 55 °C)Thermophilic operation appeared to offer no advantage over the moreconventional mesophilic temperatures. Tables 56, 57, 58 and 59 compare the meansulphate reduction efficiencies, the mean total dissolved suiphide concentrations, the meanTOC removal efficiencies and the mean gas production rates for the common levels ofinvestigation of the three sulphur management strategies.a) SulphateMesophilic temperatures resulted in approximately double the sulphate removalefficiencies compared to the 55 °C thermophilic experiments which investigated the effectsof acid phase reactor pH. A summary of these results is shown in Table 56. (It should benoted that the sulphate removal efficiencies presented for the methane phase reactors areoverall efficiencies which include the sulphate removed in the acid phase reactors.) Thisgeneral observation does not apply to the lowest HRT reactors where the differences wereusually much smaller. Rintala et al. (1991) also recorded low sulphate reduction efficiencyof 24 to 64% in UASB reactors treating a sulphate rich TMP effluent at 55 °C. Theyascribed these incomplete sulphate removals to substrate limitations. However, the high189VFA levels which were measured in these experiments do not support such anexplanation.While the sulphate reduction efficiencies were inhibited down to comparable levelsat 1.0 mM molybdate for both of the temperature regimes, the change in the thermophilicsulphate reduction was not as marked because of the comparatively small extent ofsulphate reduction (just 40 to 50%) for the zero molybdate baseline. Similar to the resultsof the 35 °C investigation, 55 °C scrubbing and stripping of hydrogen suiphide had nosignificant effect on sulphate reduction. Overall, mesophilic operation exhibited muchgreater sulphate reductions.Table 56: Effect of Sulphur Management Strategies and Reactor Temperature onSulphate ReductionMean Sulphate Reduction Efficiency [%]35°C 55°CAcid/Methane Phase Acid/Methane PhaseSulphur Reactor ERT [daysi Reactor HRT [days]Management 0.2/0.4 0.4/0.8 0.6/1.2 0.2/0.4 0.4/0.8 0.6/1.2Acid PhaseReactor pH:7.0 40/49 82/89 77/87 36/46 33/48 19/427.5 47/79 78/86 83/88 32/40 36/43 34/508.0 39/46 78/84 84/85 25/33 19/39 22/51MolybdateAddition [mM]:0 35/59 56/90 58/89 32/40 36/43 34/500.5 49/57 66/91 56/93 23/34 26/36 20/371.0 17/26 22/40 15/40 19/30 20/33 17/33GasStripping:water scrubber control 7 / 30 41 / 87 28 / 85 23 / 36 24 / 38 5 / 22femcchloridescrubber 38/56 43/72 47/83 32/34 34/42 10/27b) Total Dissolved SuiphideAlthough the sulphate removal efficiencies were generally much higher in themesophilic operations compared to the thermophilic runs (Table 57), this was not always190reflected in the total dissolved concentrations. This phenomenon may have been due toeffluent variability. It may also have arisen from the temperature effects on the microbes oron the chemical properties of the reactor liquor. Since the Henry’s Law absorptioncoefficient for hydrogen sulphide is lower at 55 °C than at 35 °C (1.59 versus 1.83), thegas phase can be expected to be richer- in-hydrogen sulphide at the- thermophilic-temperatures with corresponding lower dissolved sulphide levels.Table 57: Effect of Sulphur Management Strategies and Reactor Temperature onTotal Dissolved Suiphide ConcentrationMean Total Dissolved Suiphide Concentration [mg/Il35°C 55°CAcid/Methane Phase Acid/Methane PhaseSulphur Reactor HRT [days] Reactor HRT [days]Management 0.2/0.4 0.4/0.8 0.6/1.2 0.2/0.4 0.4/0.8 0.6/1.2Acid PhaseReactor pH:7.0 25/40 82/87 67/77 91/113 80/112 58/1167.5 40/59 74/79 76/73 42/61 48/61 50/838.0 46/59 112/105 102/101 59/85 62/109 62/131MolybdateAddition [mM]:0 38/67 62/95 68/97 42/61 48/62 50/830.5 56/69 85/111 75/110 25/60 37/63 29/631.0 4/11 20/30 9/27 26/49 31/47 24/55GasStripping:waterscrubbercontrol 132/186 211/264 170/266 83/125 92/124 22/71ferric chloride scrubber 69/58 103/92 99/93 22/52 95 /52 25/48c) Wastewater Treatment Efficiency:The sulphur management strategies investigated revealed that significantlyimproved treatment efficiencies are possible in spite of the wastewater COD:BOD5ratiosof 2:1 to 3:1. This would indicate a ceiling on the treatment efficiency of roughly 33 to50%. Several of the sulphur management experiments exceeded this value. However,these experiments demonstrated that a considerable fraction of the dissolved organic191compounds in this effluent was still not readily anaerobically bio-degradable under eithermesophilic or thermophilic conditions or with any of the sulphur management strategies.Judging from the high VFA levels which were measured both in the feed and in thereactors (Tables 59 to 63), inhibition of the acetoclastic MPB or SRB appears to be thecause of the poor treatment efficiency rather than a recalcitrance per se of wastewaterconstituents towards bio-degradation. This is in contrast to the conclusions of Potapenkoet al. (1993) who attributed incomplete oxidation of carbohydrates to the pooraccessibility, for the acid forming bacteria, of some carbohydrates bound in lignincarbohydrate complexes.A comparison of the wastewater treatment efficiencies at 35 °C and 55 °C for thesulphur management strategies is presented in Table 58. The TOC removal efficiencies forthe methane phase reactors are overall efficiencies and they include the TOC which wasremoved in the acid phase reactors.Zehnder et al. (1982) and Isa et al. (1986) found that both sulphate reduction andmethane formation could occur simultaneously at concentrations up to 1000 mg/isulphate. When sulphate concentrations were not limiting, the work of Lovley et aL(1982) demonstrated that the SRB inhibited methane production by scavenging hydrogenand lowering the hydrogen concentration to below a minimum level necessary forutilization by the MPB. Consequently, they surmised that the lack of methane productionin high sulphate effluents was not due to a significant toxic effect.In order to enhance the treatment efficiency of this wastewater with its highconcentrations of wood extractives, MacLean et aL (1990) diluted the Quesnel River Pulpeffluent stream with clarified effluent from their aerated stabilization basins at dilutionlevels up to 60%. MacLean et al. (1990) reported 45 and 55% COD and BOD5 removalefficiencies respectively of a mesophilic two phase pilot plant treatment system withloading rates from 9 to 18.3 kg COD/m3day. They followed the anaerobic stage with a1.1 day HRT activated sludge reactor train to result in overall 60% COD and 90% BOD5192removal efficiencies and complete toxicity removal. Rankin et a!. (1992) found that aUASB pre-treatment could reduce the aeration time in a single stage activated sludge unitfrom 64 to 48 hours to produce a low BOD, low SS and non-lethal discharge.Unlike the mesophilic studies, changing the acid phase reactor pH from 7.0 to 8.0at 55 °C did not appreciably enhance the TOC removal efficiency. Peak TOC removaloccurred at a total 1.8 day HRT at an acid phase reactor pH of 7.5 for both temperatures.Under these conditions, 63% TOC was removed at 35 °C whereas only 24% was removedat 55 °C.At 35 °C, molybdate addition to 1.0 mlvi resulted in lower treatment efficiencieswhereas 1.0 mM molybdate addition to the feedstock resulted in increased treatmentefficiencies at thermophilic conditions. Puhakka et at. (1988) were convinced from theirstudies that molybdate inhibition of SRB was not feasible due to its adverse effects on theMPB.Gas scrubbing and stripping of hydrogen suiphide from the MPR boosted the TOCremoval efficiency for both temperature regimes. Scrubbing and stripping at 35 °C resultedin approximately double the TOC removal efficiency compared to the control run. Hiltonand Oleszkiewicz (1987) also observed the wastewater treatment efficiency to bepositively correlated with the reduction of sulphate, in particular for suiphide strippedreactors. 55 °C scrubbing also improved the TOC removal efficiency from a baselinecomparable to the 35 °C operation but the increase was small.Unlike these results, other investigators have achieved high removal efficienciesusing thermophilic treatment. Salkinoja-Salonen et aL (1983) reported improved treatmentefficiency of TMP effluent at 60 °C compared to 35 °C on a pilot scale. Their carbonremovals were not reflected in the methane gas production rates which were highlyvariable and fractions of what were anticipated from theory.193Table 58: Effect of Sulphur Management Strategies and Reactor Temperature onTotal Organic Carbon RemovalMean TOC Removal Efficiency [%135°C 55°CSulphur HRT jdaysl HRT [dayslManagement 0.6 1.2 1.8 0.6 1.2 1.8Acid PhaseReactor pH:7.0 11/14 7/26 14/23 18/22 14/17 9/117.5 20/24 39/55 36/63 13 /20 16/18 19/248.0 11/23 24/41 26/39 4/19 10/13 8/20MolybdateAddition [mM]:0 22/29 22/25 27/34 13/20 16/18 19/230.5 31/38 33/41 33/42 15/19 19/21 15/171.0 22/22 15/36 24/29 20/29 17/29 23/32GasStripping:water scrubbercontrol 5/22 22/24 17/24 27/27 21/28 7/32ferric chloride scrubber 29/43 38/51 41/56 27/37 34/39 26/36d) Gas Production Rate:The gas production rate and methane content of the gas were monitoredthroughout the experiment program. A summary of the gas production rates is given inTable 59.Gas production was consistently greater at the longer HRTs. The increase of gasproduction with HRT suggests that the treatment efficiency could benefit by furtherincreased reaction times. Thermophilic gas production rates were much lower, in the orderof approximately 10%, than the mesophilic experiments. However, the TOC removalswere no more than 50% smaller for the thermophilic runs than for the mesophilicoperation. The mechanism of TOC removal is therefore not only by gas production. SinceTOC assays were performed on centrifuged samples, it is likely that a large portion of theorganic constituents of the effluent were removed with the solids. Biomass growth andprecipitation of organic compounds are the probable routes for organics removal from the194liquid phase. It should be noted that the VFAs are not appreciably volatile under the nearneutral reactor pH conditions of these experiments.Table 59: Effect of Sulphur Management Strategies and Reactor Temperature onMean Total Gas Production RateMean Gas Production Rate jlldayj35°C 55°CSulphur HRT [daysi HRT [daysiManagement 0.6 1.2 1.8 0.6 1.2 1.8Acid PhaseReactor pH:7.0 1.4 3.5 3.3 0.1 0.2 0.27.5 1.6 3.5 7.3 0.3 0.1 0.18.0 1.1 2.2 3.1 0.1 0.1 0.2MolybdateAddition [mM]:0 0.9 1.5 1.2 0.1 0.12 0.300.5 0.9 1.6 2.4 0.08 0.14 0.271.0 0.1 0.7 0.8 0.12 0.24 0.36GasStripping:water scrubber control 0.3 0.6 1.1 0.1 0.1 0.3ferric chloride scrubber 0.3 1.5 1.2 0.2 0.2 0.5e) Volatile Fatty Acids:Acetate: High acetate levels leaving the reactor system under low organic loadingconditions indicates that either the MPB were inhibited or that the MPB consortia did notinclude methane forming species which were capable of degrading acetate. Table 60summarizes the acetate concentrations for the sulphur management strategies at bothoperating temperatures. While there may have been competition between the SRB andIVIPB for available hydrogen, clearly there is not a competitive relationship between thebacterial groups for acetate.In highly loaded, suiphide producing reactors, Hilton and Oleszkiewicz (1987) alsofound sulphide to be antagonistic towards the acetoclastic MPB or SRB. In a stabledigester, acetate is consumed by the MPB and the SRB to keep concentrations low.195McFarland and Jewell (1990) measured large accumulations of acetate at influent sulphatelevels exceeding 225 mg/I sulphate. At a 10 day HRT and 10 g COD/I feed, theydiscovered a linear relationship between the acetate and the influent sulphate levels. Fromapproximately 600 mg/l acetate with no added sulphate, acetate accumulated to greaterthan 3000 mg/I at 1 500-mg/i sulphate. Their response-was. similar-to what was observed-in-.. .- -these experiments.Mulder (1984), Ueki et at. (1986), Isa et a!. (1986) also observed the contributionof acetate as an electron donor for sulphate reduction to be insignificant. This wasmanifested by lower sulphate reduction efficiency and decreased hydrogen suiphide gasconcentrations. The reason for this lack of acetate uptake is unknown. Mulder (1984)ascribed this phenomenon to exceeding a critical COD:S04ratio.The SRB which oxidize acetate, the Desz4fobacter, are reportedly widespread inmarine and brackish water sediments (Widdel and Pfenning, 1977). Winfrey and Ward(1983) found acetate to be a major electron donor for sulphate reduction in marinesediments. Perhaps such bacteria cannot thrive in low salinity anaerobic digesterenvironments. A number of authors (Zeikus and Wolfe, 1972, Daniels eta!. 1977, Zeikus,1980) have not included acetate among the substrates which can fhnction as an energysource for Methanobacterium thermoautotrophicum. On the other hand, other authorshave recorded high treatment efficiencies at thermophilic temperatures (Rintala andLepisto, 1992, 1993). It would appear that, similar to the mesophilic methanogens, somethermophilic methanogens can degrade acetate and others cannot.The apparent lack of a symbiotic relationship between the acetate generating SRBand non-SRB acetogens and the methane producing bacteria appears to be due toinhibition of the MPB. Puhakka et a!. (1989) found a decrease in the viable counts ofacetate utilizing MPB after 43 days of incubation in a simulated evaporator condensatewastewater which contained sulphate, sulphite or dithionite. After stripping H2S to below196a toxic level, Puhakka et al. (1989) concluded that the inhibition of the MPB was due tothe initial oxidation state of sulphur rather than the resultant suiphide concentrations.Undegraded acetate is consistent with the findings of Rintala et al. (1991) whoanaerobically treated sulphate rich TMP effluent in UASB reactors at thermophilictemperatures. The low methane gas production rates which were observedthroughout thexperimental program also confirmed an inability to utilize acetate.Formate, butyrate, propionate and lactate were also measured for theseexperiments. Briefly, the effects of the sulphur management strategies on formate,butyrate, propionate and lactate are as follows.Formic Acid: Formate concentrations in the feed were small and were decreased to nearzero levels in the methane phase reactors for the three HRTs at both mesophilic andthermophilic temperatures (see Table 61). With formate not utilized by the SRB (Isa et al.,1986), its removal indicates some methanogenic activity. Molybdate, added at levels whichinhibited the SRB and MPB, led to non-zero formate levels exiting the reactors.n-Butyric Acid: More than the other VFAs, fluctuating influent concentrations confoundthe ability to draw conclusions of the sulphur management strategies on butyrate levels.The mean butyrate concentrations are summarized in Table 62. As the APR pH wasincreased from 6.5 to 8.0, the butyrate concentrations diminished. Increased influentsulphate concentrations also led to lowered butyrate levels in the reactors. Strippingsulphide from solution appeared to exert no effect on butyrate concentrations.Propionic Acid: These experiments demonstrated the influence of the sulphurmanagement strategies on propionic acid levels. Table 63 summarizes these trends. Inspite of the high variability of the influent propionate concentrations, in general it can beconcluded that the net consumption of propionate was promoted by suiphide stripping andby elevated sulphate concentrations. Ueki et aL (1986) also observed propionate uptake tobe strongly promoted by the addition of sulphate. In serum bottle tests, McCartney andOleszkiewicz (1990) measured propionate accumulations at elevated un-ionized suiphide197concentrations. Molybdate addition, at levels which inhibited the SRB, led to a netgeneration of propionic acid in these experiments (Table 62).Lactate: Lactate, the main source of energy for the SRB, is readily removed in anaerobicsystems without sulphate limitations (Cappenberg and Prins, 1974). The mean lactic acidconcentrations for all of the experiments are summarized in Table 64. For all but themolybdate experiments, lactate was consumed to near zero levels in the MPRs. The smallresiduals which remained suggest that the SRB were not substrate limited. Inhibition ofthe SRB resulted in non-zero lactate concentrations.An increase in the acid phase reactor pH led to lower butyrate, propionate andlactate levels, indicative of active SRB. Increased molybdate concentrations lead toincreased VFA levels since the SRB and the MPB were inhibited. The gas scrubbingexperiments resulted in the lowest VFA concentrations, perhaps due to alleviated sulphideinhibition of the SRB and MPB. Thermophilic temperatures lead to VFA concentrationsgreater than for the comparable mesophilic experiments.198Table60:EffectofSulphurManagementStrategiesandReactorTemperatureonAceticAcidConcentrationferricchloridescrubberbatch2+sulphateMeanAceticAcidConcentration[mg/i]±5%CV35°C55°CSulphurAcid/MethanePhaseReactorHRT[days]Acid/MethanePhaseReactorHRT[days]ManagementFeed0.2/0.40.4/0.80.6/1.2Feed0.2/0.40.4/0.80.6/1.2AcidPhaseReactorpH:6.57.07.5 8.0 MolybdateAddition[mM]:0batch10.1batch10.5batchl0.75batch11.0batch11.0batch21.0batch3GasStripping:controlferricchloridescrubberbatch1ferricchloridescrubberbatch211601180/18741655/15391730/16179151455/17401604/17341513/17787051155/11971075/11811044/1172915825350625/545655/200465/2608801325/13751665/13201690/12809701345/14151660/10751660/5108901260/13701750/13601720/1565730910/15151410/11551240/149511601180/18751655/15401730/162011101480/13101610/14951780/12606201295/9801480/12301690/9007151050/12101425/8551155/995755480/5201110/1120785/115010101220/7101200/12151095/125513411805/18951905/19201078/177014651800/18951850/19101875/16356001030/990865/810955/7855051040/1030935/9101015/83011757706251455/17401604/17341513/17781424/15691491/16671445/16402264/24652243/24332174/26331269/15421413/15421142/14101517/18111505/12771213/1616Table61: Effectof SulphurManagement StrategiesandReactorTemperatureonFormicAcidConcentrationMeanFormicAcidConcentration[mgll]±5%CV35°C55°CSulphurAcid/MethanePhaseReactorHRT[days]Acid/MethanePhaseReactorRRT[days]ManagementFeed0.2/0.40.4/0.80.6/1.2Feed0.2/0.40.4/0.80.6/1.2AcidPhaseReactorpH:6.500/00/00/0----7.06129/230/00/025015/02/00/07.55428/120/08/03398/19/114/18.06816/138/08/01536/04/024/0MolybdateAddition[mM]:Obatchi839/0.4/06/03408/19/114/10.lbatchl25016/00/20/0----0.Sbatchl2000/00/06/32342/06/08/00.75batchl22621/038/00/0----1.Obatchl7739/1117/415/026336/012/015/01.Obatch212164/7512/40/2---1.Obatch313620/82/24/2--GasStripping:control813/30/00/02113/010/581/0ferricchloridescrubber1560/00/03/01287/59/429/0batch1ferricchloridescrubber900/00/03/0batch2ferricchloridescrubber780/04/03/0batch2+sulphateTable62:EffectofSulphurManagementStrategiesandReactorTemperatureonn-ButyricAcidConcentrationCMeann-ButyricAcidConcentration[mg/I]±5%CV35°C55°CSulphurAcid/MethanePhaseReactorHRT[daysiAcid/MethanePhaseReactorHRT[days]ManagementFeed0.2I0.40.4/0.80.6I1.2Feed0.2/0.40.4/0.80.6I1.2AcidPhaseReactorpH:6.566226/792358/917341/292----7.098164/133385/46111/015531/7780/13090/907.54953/4722/2514/02371/8034/4665/418.054112/8288/8013/06343/2480/7427/26MolybdateAddition[mM]:Obatchi7734/4040/5056/592371/9034/4665/410.lbatchl15432/7683/12689/94----0.5batchl8862/5370/6060/585281/6349/4748/410.75batchl7695/1616/010/0----1.Obatchl6672/7343/3335/3911444/5331/2745/491.Obatch27033/8613/2485/73----1.Obatch300/1616/024/0----GasStripping:control2513/90/40/05355/3945/35175/0ferricchloridescrubber3311/08/03/52025/1517/14146/112batch1ferricchloridescrubber6030/3118/023/22----batch2ferricchloridescrubber6512/87/00/0----batch2+sulphateTable63:Effectof SulphurManagementStrategiesandReactorTemperatureonPropionicAcidConcentrationMeanPropionicAcidConcentration[mg/I]±5%CV35°C55°CSulphurAcid/MethanePhaseReactorHRT[days]Acid/MethanePhaseReactorRRT[days]ManagementFeed0.2I0.40.4/0.80.6I1.2Feed0.2I0.40.4I0.80.6/1.2CAcidPhaseReactorpH:6.57.07.5 8.0 MolybdateAddition[mM]:0batch10.1batch10.5batch10.75batch11.0batch11.0batch21.0batch3GasStripping:controlferricchloridescrubberbatch1ferricchloridescrubberbatch2ferricchloridescrubberbatch2+sulphate18089/8078/8155/62129251/180203/173154/1356217/3642/3349/3946128/122117/61140/77137288/253312/214331/14022155/136157/102125/88112243/192230/90249/15485141/9099/7875/5718387/8277/8555/61319163/460227/19637/52137246/159132/5699/7368183/221131/14465/1098870/93105/80155/803973/9113/1730/02887/7556/5840/018298/138103/113105/35193/196171/176179/17112916771251/180203/173154/135157/215155/209132/175177/157131/9686/92128196/235179/209169/21572197/144152/13178/17031 4174/2943/049/0Table64:EffectofSulphurManagementStrategiesandReactorTemperatureonLacticAcidConcentrationMeanLacticAcidConcentration[mg/I]±5%CV35°C55°CSulphurAcid/MethanePhaseReactorHRT[days]Acid/MethanePhaseReactorJIRT[days]ManagementFeed0.2/0.40.4/0.80.6/1.2Feed0.2/0.40.4/0.80.6I1.2CAcidPhaseReactorpH:6.57.07.5 8.0 MolybdateAddition[mMJ:0batch10.lbatch10.5batch10.75batch11.0batch11.0batch21.0batch3GasStripping:controlferricchloridescrubberbatch1ferricchloridescrubberbatch2ferricchloridescrubberbatch2+sulphate610/00/015/1311161/4170/2398/2810669/6150/3034/010038/3656/4070/416852/5063/6935/3810184/6857/3760/197157/5740/2552/258257/030/5060/4011747/4672/2768/479164/5157/6162/6031148/3832/3916/1527887/7470/8978/7231377/6165/2277/3031123983/6957/3760/19276115/87110/86147/9410027/2238/2544/28276115/87110/86147/9422184/8085/74139/75297113/84120/86156/96372156/99196/103254/1509882/8170/5691/690/00/014/0420/014/013/010.3 Start-up After 40 DaysWhile the start up time of a reactor is only a small fraction of its lifetime, the timeto bring a new treatment plant into operation can be a determining factor in its selection.The recovery from upsets can also present a major expense to the mill and the timerequired to achieve efficient, steady state performance is crucial since mill shut down maybe necessary upon failure to meet discharge regulations.Figure 10 follows the performance of the reactor system for 30 days following a40 day shut down. Sulphate, sulphide and TOC concentrations were measured almostevery day for the duration of this run.Sulphate concentrations were reduced to quasi steady state values approximately10 days into the resumption of continuous feeding (Figure 10). The methane phasereactors reached their baseline concentrations of 100 to 200 mgIl sulphate within 5 days ofre-start for the two longer HRTs whereas the shorter 0.4 day methane phase reactor andeach of the acid phase reactors took approximately 10 days to stabilize. It appears that thetime to reach steady state varied inversely with the HRT. This may be due to an excessmetabolic capacity in the longer HRT reactors.The sulphide concentrations were a reflection of the sulphate trends, similar to theperformance of the sulphur management experiments. These values are graphed in Figure10. After approximately 12 days into the resumption of continuous feeding, a quasi steadystate performance was approached.Steady TOC removal was realized after only approximately 3 days after restart(see Figure 10). Steady state for the TOC removal was characterized by considerable dayto day fluctuations. Kennedy et at. (1991a) also restarted UASB reactors to achievesteady performance within 2 to 3 days to pre-shutdown levels following a week ofdormancy.2041000-900- •S_• _S-SS S N/ •800-_ 700- .-O..600- aO-O ti—4.500E8o. /A/5O 4 Q-•-O- (O9- 40- / / b- •. d --10 •.oH 02500.2000-r -1500-1000•--AI I I I0 5 10 15 20 25Start-up Time [days]_______________—•—Feed0 0.2dHRTAPRI0.4dHRTAPR20—- O,6dHRTAPR3—.—0.4dHRTMPR I—A— 0.8 d HRT MPR 2Figure 10: Start-up of the Two Phase System at 35°C:S042,S2 and TOC vs Time and HRT20510.4 Effect of HRTAs the total HRT increased from 0.6 to 1.8 days, the treatment efficiency, sulphatereduction and dissolved suiphide concentration increased for all of the experiments at 35°C. In most cases, there was no significant difference in performance between the 1.2 and1.8 day HRT settings, suggesting that there is little advantage of employing the longerHRT. Lo et al. (1991) also reported TOC removals from a BCTMP effluent to increase asthe HRT was increased from 0.5 to 2 days with further HRT increases ineffective. Thereappears to be no such relationship between HRT and TOC removal at 55 °C where all ofthe TOC removals were poor and unaffected by HRT (Table 58). Longer HRTs promotedthermophilic sulphate reduction and a corresponding sulphide formation. These generaltrends are evident upon examination of Tables 56, 57, 58 and 59.The optimum loading rate of 15 kgCOD/m3daywhich was determined by Hall etaL (1986) for NSSC wastewater was greater than what could be described as the mosteffective loading rate for these experiments. Results from the range of loading rates whichwere applied in these experiments of approximately 4 to 18 kg COD/m3day demonstratedthat the optimum was between 4 and 11 kg COD/m3day.Previous studies on the anaerobic treatment of CTMP effluents have demonstratedfairly similar reductions for the HRTs which were employed for these experiments.Andersson et at. (1985) found reductions of 62% and 90% for COD and BOD5respectively for 35 °C anaerobic treatment of a peroxide brightened CTMP effluent. Theyalso observed complete reduction of sulphate to suiphide at 10 and 36 hour HRTs for theacidogenic and methanogenic stages. They concluded that anaerobic removal of organiccompounds followed by aerobic detoxification would be the most effective configurationwith respect to total costs and treatment results. Wilson et aL (1987) recorded BOD5removals of up to 63% for anaerobic treatment systems operated at a one day HRT.However, the toxicity removal was minor and the effluent required 3 to 5 days of aerationin a polishing basin to be non-lethal. Pichon et aL (1986) compared the anaerobic206treatment of CTMP effluents which were generated from hardwood and softwoodfurnishes. HRTs of 12 and 60 hours were needed to achieve 60% COD removal and 80%BOD5 removal for the hardwood and softwood effluents respectively. They concludedthat the toxicity of the sulphur compounds in the softwood effluent was responsible forthis relative inefficiency.10.5 Effect of Phase SeparationThe total working volume of the smallest, two phase, reactor set equaled thevolume of the largest, acid phase reactor. If the performance of the reactors is dominatedby the effect of HRT, the effectiveness of phase separation for this wastewater at the 0.6day HRT could be determined simply by comparing the performances of these reactors.However, there were a number of differences between the acid phase reactor and the twophase reactor set. The acid phase reactor was operated as a UASB reactor with pHcontrol and liquid recycle whereas the methane phase reactor had no direct pH control andmore closely resembled a plug flow system. In addition, the means of biomass retentionwas different for the two reactor systems. Biomass was retained in the acid phase reactorby means of settling flocs. In the methane phase reactor, microbes were retained as anattached bioflim as well as flocs residing in the void spaces of the stationary film support.A summary of the target 0.6 day HRT acid phase reactor as well as two phasesystem experiments which manipulated the acid phase reactor pH and molybdate additionto the feedstock is compiled in Table 65. Since the gas scrubbing experiments recycledbiogas through the methane phase reactors and not through the acid phase reactors, noevaluation of phase separation could be made for these experiments.In terms of TOC removal, significant differences were observed between a singlephase reactor and a two phase reactor for these experiments, but no general trends couldbe distinguished. The trends for sulphate reduction and sulphide concentration also werenot clear. Only the mesophilic runs which examined the effects of molybdate addition207demonstrated comparable performances between the acid phase reactor and the two phasesystem at equal HRTs. The acid phase reactor pH experiments at both temperatures andthe 55 °C molybdate experiments revealed very different sulphate reductions and suiphideconcentrations for the acid phase versus the two phase system. These differences may beascribed to the differences in pH control, mixing or biomass of the acid phase reactorversus the two phase reactor.The lack of similarity of the reactor systems makes the formation of more generalconclusions concerning the effects of phase separation difficult. While the TOC removalrates were not affected by an attempted phase separation for these experiments,establishing a well mixed reactor upstream of an unmixed reactor offers the opportunitiesfor the control of pH, toxicity and biomass.208CTable65:EffectofPhaseSeparationonTOCRemoval,SulphateReductionandSulphideConcentrationfor a0.6dayHRTat35°Cand55°C35°C55°CSulphurAcidPhaseReactorTwoPhaseReactorAcidPhaseReactorTwoPhaseReactorManagementTOCSO4S2TOCSO4S2TOCSO4S2TOCSO42%<%<[mgllj%<%<FmWi]%<%<jmgfl]%<%<[mgi]AcidPhaseReactorpH:7.01477671449409195822461137.53683762479591934502040618.0268410223465982262193385MolybdateAddition[mM]:02758682259671934502039610.53356753857691520291934601.02415922261123172429304910.6 Reproducibility of ResultsThe often large, batch to batch variation in the effluent which was received fromthe Quesnel River Pulp Co. BCTMP/TMP mill tempers one’s confidence in makingcomparisons and drawing conclusions from the tests which used different effluent batches.Instead, comparisons can be drawn from the results of experiments which were performedusing a common effluent batch. Some estimation of the efficacy of a particular set ofreactor conditions can be made by measuring the difference of the performance over thatof the control setting. These differences can be compared (with some skepticism) for thedifferent sulphur management strategies which were tested using different effluent batches.The variability between effluent batches is shown in Table 66. This Table comparesthe three control runs which were performed for the three sulphur management strategies(the acid phase reactor pH, “APRpH”, molybdate addition, “Mo” and scrubbing andstripping the gas, “Scrub”) at an acid phase reactor pH of 7.5 and 35 °C.Variability was measured in terms of sulphate reduction efficiency, sulphideconcentration, TOC removal efficiency and gas production rate. It was manifested by abroad range of sulphate reduction efficiencies, sulphide concentrations, TOC removalefficiencies and gas production rates when the control runs were repeated throughout theexperiment program. These large differences indicate that the effluent batch and perhapsthe experiment day have a large effect on the measured results. While the trends derivedfrom the extremely large differences between sets of results can be reported upon, themore subtle effects are not easily interpretable.210Table66: ReproducibilityofResults:RepeatsofAPRpH7.5ControlsThroughouttheExperimentsat35°CLJacidphasereactorpHexperimentsetmolybdateadditionexperimentsetscrubbingandstrippinghydrogensuiphidefromthemethanephasereactorexperimentsetHRTandReactorS042RemovalS2TOCRemovalGasProductionRate1%](mg/I](%](I/day]APRpHMoScrubAPRpHMoScrubAPRpHMoScrubAPRpHMoScrub0.2dHRTAPR1473574038132202215---0.4dHRTMPR179593059671862429221.60.90.30.4dHRTAPR27856417462211392222---0.SdHRTMPR286908779952645525243.51.50.64O.6dHRTAPR38358287668170362717---1.2dHRTMPR388898573972666334247.31.21.1Legend:APRpHMoScrubChapter 11Conclusions1) Manipulating the concentrations of sulphur in the effluent resulted in significantlyimproved treatment efficiencies. This would decrease the aeration, nutrient and processingtime requirements in a downstream biological aerobic treatment system as well as decreasethe disposal problem of excess activated sludge.2) Changing the acid phase reactor pH promoted TOC removal but did not address theinability of the methane producing bacteria to degrade acetate and to produce methane.Therefore free hydrogen sulphide was not the only compound which was inhibiting themethanogens.3) Adding molybdate at 1.0 mM to the effluent decreased the extent of sulphate reductionand resulted in markedly lower dissolved suiphide concentrations effectively in two out ofthree cases. The methanogens however did not fill the niche left by the inhibited sulphurreducing bacteria and consume organic compounds. Therefore, molybdate, sulphate orother compounds inhibited the methanogens,4) Stripping dissolved sulphide from the methane phase reactors promoted wastewatertreatment efficiency to result in the highest TOC removals of any of the sulphurmanagement strategies. In spite of the promoted organic removal activity, very highacetate levels persisted in the reactors, indicating inactive acetoclastic MPB or SRB.Therefore, very small levels of sulphide or sulphate or other compounds in the effluentinhibited the methane forming bacteria.5) Increasing the reactor temperature from the 35 °C mesophilic regime to thermophilic(55 °C) temperatures significantly worsened treatment efficiencies. This could be due tothe fewer number of microbial species which function at these temperatures or to lowerthermophilic activity compared to mesophilic activity. Cooling the effluent to 35 °C iswarranted prior to anaerobic treatment.2126) Increasing the total FIRT from 0.6 to 1.8 days resulted in increased biochemicalconversions of sulphur compounds, TOC and VFAs to form acetate. A HRT of 0.6 dayswas less effective than the 1.2 and 1.8 day HRT reactors. There was little difference in theperformance of the two longer HRT reactors. Although significantly extending the F1RTwould lead to greater substrate conversions, there appears to be little practical benefit intreating the effluent for longer than 1.2 days. Since the microbes in the treatment systemwere inhibited throughout the experimental period, an optimum HRT cannot be estimated.7) While the sulphur management strategies of pH adjustment, inhibition of the sulphurreducing bacteria and stripping of the dissolved suiphide from the methane phase reactorswere effective in manipulating the fate of the inorganic sulphur species, this did not resultin even remotely satisfactory treatment efficiencies such that the treated effluent could bedischarged directly to receiving waters. The high levels of acetic acid which were observedthroughout these experiments suggest that compounds in addition to inorganic sulphurinhibited the treatment microorganisms.213Chapter 12Recommendations for Further StudySulphur management significantly improved the anaerobic treatment efficiency ofthe BCTMP/TMP effluent; However, compounds in addition to sulphur appear to beresponsible for inhibiting the methanogens. Therefore, continuing to address the sulphurproblem should be pursued while minimizing the exposure to the suspected additionalinhibitory agents.1) Discontinue the molybdate investigation. Inhibiting the SRB led to greatly decreasedsulphide levels but the molybdate levels which were required to achieve this also began toinhibit the MPB. Thus, no trend towards markedly enhanced TOC removal efficiencieswas observed. Molybdate addition does not appear to be a viable option for effectivelytreating this mechanical pulping effluent so this process is unlikely to be worth futureinvestigation.2) Allow the SRB to function effectively in the acid phase reactor to completely reducethe inorganic sulphur compounds to sulphide and consume organic compounds in theprocess. Strip the dissolved suiphide after the acid phase reactor and prior to theintroduction of effluent into the methane phase reactor. This would protect the MPB fromthe inhibitory effects of dissolved sulphide and preclude the precipitation of trace metalmicronutrients as metal suiphides. An independent gas stripper outside of the bioreactorsmay be operated at greater efficiencies than a combined stripper/MPR or a stripper/APR.A schematic diagram of the recommended treatment system is given in Figure 11.3) Add nutrients both before and after sulphur reduction, that is, to both the acid phaseand methane phase reactors. The precipitation of the trace metal micronutrients asinsoluble sulphides could compromise the microbial activity in the acid phase reactor. It isprobable that these metals will be precipitated as metal sulphides and consequently they214Figure 11: Recommended Treatment System215to Solids Disposalwill be unavailable to the downstream methane forming bacteria. Also, since largeconcentrations of uncomplexed DTPA are present in the effluent, add micronutrients asperiodic slug doses at concentrations which exceed the sum of the precipitating potentialof the suiphide and the complexing capacity of the chelating agent.4) Since wood extractives which-are known to inhibit the methanogens and resin acids inparticular are present at levels which may pose difficulties in treating the effluent,measures to address this should be investigated. Alternatives include dilution of theeffluent with other, less toxic streams or the precipitation of the resin acids by metalcations.5) Thermophilic anaerobic treatment remains attractive but inconsistently proven for thistype of effluent. Alternative inoculation strategies could be investigated.6) While the unmixed, UASB/fixed film methane phase reactor was selected in order tomaximize the biomass density, it also precluded the ability to easily sample biomass fromthe system. A UASB with liquid recycle and independent pH control is better defined andmore easily interpreted and could circumvent some of these difficulties.216List of AcronymsADTPD air dried tons per dayAPB acid producing bacteriaAPR acid phase reactorBCTMP bleached chemithermomechanical pulpBOD5 5 day biochemical oxygen demandCOD chemical oxygen demandCODr chemical oxygen demand removedCSTR continuously stirred tank reactorCTMP chemithermomechanical pulpCV coefficient of variationDSFF down flow stationary fixed filmDTPA diethylene triamine penta acetic acidEC5O effective concentration for 50% inhibition of Microtox bacteriaHRT hydraulic retention timeHRTT total hydraulic retention timeMPB methane producing bacteriaMPR methane phase reactorNSSC neutral suiphite semi-chemicalp probabilityr pearson correlation coefficientsquared multiple rSD standard deviationSRB sulphur reducing bacteriaSRT solids retention timeSTEL short term exposure limitTLV threshold limit valueTOC total organic carbonTMP thermomechanical pulpTSS total suspended solidsUASB upflow anaerobic sludge bedVFA volatile fatty acidVS volatile solidsVSS volatile suspended solids50%IC concentration causing 50% methanogenic inhibition96hLC5O 96 hour median lethal concentration217ReferencesAbram, J.W. and Nedwell, D.B., (1978). Inhibition of methanogenesis by sulphate reducing bacteriacompeting for transferred hydrogen. Archives ofMicrobiology, 117, 89-92.ACGIH, (1988). Documentation of the threshold limit values. American Conference of GovernmentalIndusthal Hygienists, 4th edition, Cincinnati, Ohio.Aitken, M.D. and Mullennix, R.W., (1992). Another look at thermophilic anaerobic digestion ofwastewater sludge. Water Environment Research, 64, 915-919.Andersson, P.E., Gunnarsson, L., Oisson, G., Welander, T. and Wikstrom, A., (1985). Anaerobictreatment of CTMP effluent. TAPPI Environmental Conference, 11-16.Andersson, P.E., Gunnarsson, L., Welander, T. and Wikström, A., (1986). Detoxification of CTMPeffluent to enable anaerobic treatment. Proceedings of the Third International Conference onBiotechnology in the Pulp and Paper Industry, Stockholm, June 16-19, 207-208.Andrews, J.F. and Pearson, E.A., (1965). Kinetics and characteristics of volatile acid production inanaerobic fermentation processes. International Journal ofAir and Water Pollution, 9, 439-46 1.APHA, (1989). Standard Methods for the examination of water and wastewater. 17th edition, AmericanPublic Health Association, Washington, D.C., USA.Archer, D.B., (1983). The microbiological basis of process control in methanogemc fermentation ofsoluble wastes, Enzyme and Microbial Technology, 5, 162-170.Asinari di San Marzano, C.M., (1981). Volatile fatty acids, an important state parameter for the control ofthe reliability and the productivities of methane anaerobic digestion. Biomass, 1, 47.Badziong, W. and Thauer, RK., (1978). Growth yields and growth rates of Desufovibrio vulgaris(Marburg) growing on hydrogen pius sulfate and hydrogen plus thiosulfate as the sole energy sources.Archives ofMicrobiology, 117, 209-2 14.Baich, W.E., Fox, G.E., Magrum, L.J., Woese, C.R. and Wolfe, RS., (1979). Methanogens: reevaluationof a unique biological group. Microbiological Reviews, 43, 260-296.Balch, W.E. and Wolfe, R.S., (1976). New approach to the cultivation of methanogenic bacteria: 2-mercaptoethanesulfomc acid (HS-CoM)-dependent growth of Methanobacterium ruminantium in apressurized atmosphere. Applied Environmental Microbiology, 32, 781.Bambrick, D.R., (1985). The effect of DTPA on reducing peroxide decomposition. TAPPI Journal, 68, 96-100.Banat, I.M., Lindstrom, E.B., Nedwell, D.B. and Balba, M.T., (1981). Evidence for coexistence of twodistinct functional groups of sulfate-reducing bacteria in salt marsh sediment. Applied EnvironmentalMicrobiology, 42, 985-992.Barford, J.P., Cad, R.G., Callander, I.J. and Floyd, E.J., (1986). Anaerobic digestion of high-strengthcheese whey utilizing semicontinuous digesters and chemical flocculant addition. Biotechnology andBioengineering, 28, 1601-1607.Battersby, N.S., (1988). Sulphate-reducing bacteria. Methods in Aquatic Bacteriology. 269-299.218Beak Consultants, (1986). Anaerobic treatment of TMP/CTMP wastewater. A Report for EnvironmentCanada Wastewater Technology Centre.Bennett, D.H., Reeser, D., Campbell, A.G. and Falter, C.M. (1988). Pre-treatment of CTMP effluent bylime to reduce resin acids, sulphite and toxicity, TAPPI Environmental Conference, 199-207.Bhattacharya, S.K., (1986). Toxic substances in methane fermentation systems: Fate and effects onprocess kinetics, Doctoral Thesis, Drexel University, Philadelphia, Pennsylvania.Boone, D. and Mali, R., (1987), Transitional bacteria. In: Anaerobic Digestion of Biomass, Chynoweth,D.P. and Isaacson, R. (editors), Elsevier Applied Science, London, 35-63.Biyant, M.P., (1977).Microbiology of the rumen. In: Sevensen (editor), Duke’s Physiology of domesticDomestic Animals. 9th edition, Cornell University Press, Ithica New York, 287.Bryant, M.P., (1979). Microbial methane production -- theoretical aspects. Journal ofAnimal Science, 48,193-201.Bryant, M.P., Varel, V.H., Frobish, R.A. and Isaacson, H.R, (1976). Biological potential of thermophilicmethanogenesis from cattle wastes. In: Microbial Energy Conversion, Schegel, H.Cl. and Barnes, J.,(editors), 347-359.Bryant, M.P., Campbell, L.L., Reddy, C.A. and Crabill, MR. (1977). Growth of Desz4fovibrio in lactateor ethanol media low in sulfate in association withH2-utilizing methanogenic bacteria. Applied andEnvironmental Microbiology 33, 1162-1169.Buhr, H.O. and Andrews, J.F., (1977). Review paper: the thermophilic digestion process., WaterResearch, 11, 129-143.Buisman, C.J.N., Lettinga, G., Paasschens, C.W.M. and Habets, L.H.A., (1991). Biotechnologicalsuiphide removal from effluents, Water Science Technology, 24, 347-356.Bull, M.A., Sterritt, R.M. and Lester, J.N., (1984). An evaluation of single- and separated- phaseanaerobic industrial wastewater treatment in fluidized bed reactors. Biotechnology and Bioengineering,26, 1054-1065.Butlin, K.R., Selwyn, S.C., Wakerley, D.S., (1956). Suiphide production from sulphate-enriched sewagesludge. Journal ofApplied Bacteriology, 19, 3-15.Callander, I.J. and Barford, J.P., (1983a). Precipitation, chelation, and the availability of metals asnutrients in anaerobic digestion II: applications. Biotechnology and Bioengineering, 25, 1959-1972.Callander, 1.1. and Barford, J.P., (1983b). Anaerobic digestion of high sulphate cane juice stillage in atower fermenter, Biotechnology Letters, 5, 755-760.Cappenberg, Th. E., (1972). Ecological observations on heterotrophic, methane oxidizing and sulfatereducing bacteria in a pond. Hydrobiologia, 40, 47 1-485.Capenberg, Th. E., (1975). A study of mixed continuous cultures of sulfate-reducing and methane-producing bacteria. Microbial Ecology, 2, 60-72.219Cappenberg, T.E. and Prins, R.A., (1974). Interrelations between sulfate-reducing and methane-producingbacteria in bottom deposits of a freshwater lake. III. Experiments with 14C labeled substrates. Antonie vanLeeuwenhoek, 40, 457-469.Chambers, L.A. and Tmdinger, P.A., (1979). Thiosulfate formation and associated isotope effects duringsulfite reduction by Closfridium pasteurianum. Canadian Journal ofMicrobiology, 25, 719-721.Clausen, E.C. and Shaw, A.H., (1981). Temperature effects on the kinetics of the anaerobic digestion ofcrop residues to methane, Fundamentals and Applications of Solar Energy — Part II, AIChE SymposiumSeries, 77, 78-85.Claypool, G.E. and Kaplan, I.R., (1974). The origin and distribution of methane in marine sediments. In:I.R. Kaplan (Editor), Natural Gases in Marine Sediments. Plenum Press, New York, 99-139.Cocci, A.A., Landine, R.C., Viraraghavan, T. and Brown, G.J., (1982). Anaerobic treatment of pulp andpaper wastes. Pulp and Paper Canada, 83, T253-T256.Cohen, A., (1983). Two-phase digestion of liquid and solid wastes. Proceedings of the Third InternationalSymposium on Anaerobic Digestion, August 14-19, Boston, Mass., Evans and Faulkner Inc., 123-138.Cohen, A., Breure, A.M., van Andel, J.G. and van Deursen, A., (1980). Influence of phase separation onthe anaerobic digestion of glucose - I maximum COD-turnover rate during continuous operation, WaterResearch, 14, 1439-1448.Cornacchio, L.-A. and Hall, E.R, (1988). Characteristics of CTMP wastewaters. Paper presented atSeminar on Treatment ofCTMP Effluents, 27 October, Vancouver, 1-14.Daniels, L., Fuchs, G., Thauer, RK. and Zeikus, J.G., (1977). Carbon monoxide oxidation bymethanogemc bacteria. Journal ofBacteriology, 132, 118.de Haast, J., Britz, T.J. and Novello, J.C., (1986). Effect of different neutralizing treatments on theefficiency of an anaerobic digester fed with deproteinated cheese whey. Journal of Dairy Research, 53,467-476.de Vries, J., (1989). Short term influence of DTPA on anaerobic sludge. Internal report of Paques B.V.,The Netherlands and Paques-Lavalin, Canada.Diekert, G., Weber, B. and Thauer, R.K., (1980). Nickel dependence of factor 430 content inmethanobacterium thermoautotrophicum. Archives ofMicrobiology, 127, 273-278.Dinsmore, N., (1987). Anaerobic treatment allows viable handling of bleached CTMP effluent. Pulp andPaper, October, 139-140.Dohanyos, M., Kosova, B., Zabranska, I. and Grau, P., (1985). Production and utilization of volatile fattyacids in various types of anaerobic reactors, Water Science Technology, 17, 191-205.Domka, F. and Gasiorek, J., (1975). Effect of the concentration of available carbon compounds on themicrobial reduction of sulphates. Acta Microbiologica Polonica, 7, 97-10 1.Domka, F. and Szulczynski, M., (1980), Studies on thiosulfate reduction by Desufovibrio vulgaris. ActaMicrobiologica Polonica, 29, 265-272.220Driessen, W.J.B.M. and Wasenius, C-O., (1993). Combined anaerobic /aerobic treatment of peroxidebleached TIvIP mill effluent. Proceedings 4th IA WQ Symposium on Forest Industry Wastewaters, June 8 -11, Tampere, Finland.Dubeski, C.V., (1993). Aerobic treatment of CTMP wastewater in sequencing batch reactors. Master ofApplied Science thesis, University ofBritish Columbia.Duff, S.B.J. and Kennedy, K.J., (1982). Effect of hydraulic and organic overloading on thermophilicdownflow stationary fixed film (DSFF) reactor. Biotechnology Letters, 4, 8 15-820.Eastman, J.A. and Ferguson, J.F., (1981). Solubilization of particulate organic carbon during the acidphase of anaerobic digestion. Journal Water Pollution Control Federation, 53, 3 52-366.Eeckhaut, M. and Alaerts, G., (1986). Anaerobic treatment of paper mill effluents using polyurethanecarrier reactor (PCR) technology. PIRA Conference on Cost Effective Treatment of Paper mill EffluentsUsing Anaerobic Technologies, 14-15 January, Leatherhead, England, 1-11.Endo, G. and Toya, Y., (1985). Anaerobic biological decomposition of malodorous compounds in kraftpulping wastewater. Water Science Technology, 17, 3 9-52.Environment Canada, (1984). Hydrogen Sulphide: Environmental and Technical Information for ProblemSpills. Enviromnental Protection Service.Ferguson, J.F. and Benjamin, M.M., (1985). Studies of anaerobic treatment of suffite process wastes.Water Science Technology, 17, 113-121.Field, J.A., (1989). The effect of tanrnc compounds on anaerobic wastewater treatment. Doctoral Thesis,Wagemngen Agricultural University, Wageningen, The Netherlands.Field, J.N., Leyendeckers, M.J.H., Sierra Alvarez, R., Lettinga, 0. and Habets, L.H.A., (1988). Themethanogenic toxicity of bark tannins and the anaerobic biodegradability of water soluble bark matter.Water Science Technology, 20, 219-240.Frenette, E.G., Jordan, J.A., Schneider, E. and Garvie, R., (1989). Start-up experience of full scale upflowanaerobic sludge bed for treatment of CTMP and NSSC wastewaters at Consolidated Bathurst’s Mill,Bathurst, New Brunswick. TAPPI Environmental Conference, 457464.Fromson, D.A., Chalmers, W.T. and Manolescu, D.R., (1986). Anaerobe$ make money treating BCTMPeffluent. CPPA Spring Conference, Jasper, Alberta.Frostell, B., (1985). Process control in anaerobic wastewater treatment. Water Science Technology, 17,173-189.Garber, W.F., Ohara, G.T., Colbaugh, J.E. and Raksit, S.K., (1975). Thermophilic digestion at theHyperion Treatment Plant, Journal ofthe Water Pollution Control Federation, 47, 950.Garvie, R., (1991). Anaerobic/aerobic treatment of NSSC/CTMP effluent and biogas utilization. 77thAnnual Meeting, Technical Section, CPPA, January 29-30, A32 1-A325.Geller, A.N. and Gottsching, L., (1985). Anaerobic fermentation of sulphite pulp mill effluents. WaterScience Technology, 17, 157-171.221Ghosh, S., Conrad, J.R. and Klass, D.L., (1975). Anaerobic acidogenesis of wastewater sludge, Journal ofthe Water Pollution Control Federation, 47, 30-45.Gibbons, S., Kovacs, T., Voss, R, O’Connor, B. and Donca, J., (1992). An appraisal of the effect ofbiological treatment on the environmental quality of high-yield mechanical pulping effluents, WaterResearch.Gujer, W. and Zehnder, A.J.B., (1983). Conversion processes in anaerobic digestion. Water ScienceTechnology, 15, 127-167.Gunnarsson, L., Andersson, P.-E. and Olsson, L.-E., (1989). Anaerobic treatment of CTMP effluent fromperoxide bleached softwood. TAPPI Environmental Conference, 535-541.Habets, L.H.A., (1987). Experience with full-scale and pilot-scale Biopaq UASB anaerobic treatment ofpulp, paper and board mill effluents. IPPTA, 24, 32-40.Habets, L.H.A. and Kuelissen, J.H., (1985). Anaerobic waste water treatment plant at Papier FabriekRoermond working successfully and saving expenses. TAPPI Environmental Conference, 93-96.Habets, L.H.A. and de Vegt, A.L., (1990). Anaerobic treatment of bleached TIvW and CTMP effluent inthe Biopaq UASB system. Proceedings 3rd IA WQ Symposium on Forest Industry Wastewaters, June,Tampere, Finland.Hall, E.R., (1987). Anaerobic treatment for BOD and toxicity reduction in pulp and paper wastewaters.Environment Canada Wastewater Technology Centre, 1-16.Hall, E.R and Adams, G.P., (1986). Anaerobic treatment of cheese whey. Proceedings PCA 0/MOESeminar on Anaerobic Fixed Film Digestion, November 5, Toronto, 65-80.Hall, E.R. and Cornacchio, L.-A., (1988). Anaerobic treatability of Canadian pulp and paper millwastewaters. Pulp and Paper Canada, 89, T188-T192.Hanaki, K., Matso, T., Nagase, M. and Tabata, Y., (1987). Evaluation of effectiveness of two-phaseanaerobic digestion process degrading complex substrate. Water Science Technology, 19, 311-322.Hardy, J.A. and Hamilton, W.A., (1981). The oxygen tolerance of sulfate-reducing bacteria isolated fromNorth Sea Waters, Current Microbiology, 6, 259-262.Harper, W.J. and Blaisdell, J.L., (1971). State of the Art Dairy Food Plant Wastes and Waste Treatment.Proceedings of the Second National Symposium on Food Processing Wastes, March 23-26 Denver,Colorado.Haschke, R.H. and Campbell, L.L., (1971). Thiosulfate reductase of Desuifovibrio vulgaris. Journal ofBacteriology, 106, 603.Hatchikian, E.C., (1974). On the role of menaquinone-6 in the electron transport of hydrogen: fumaratereductase system in the strict anaerobe Desulfovibrio gigas. Journal ofGeneral Microbiology, 81, 261.Heinrichs, D.M., Poggi-Varaldo, H.M. and Oleszkiewicz, J.A., (1990). Effects of ammonia on anaerobicdigestion of simple organic substrates. Journal ofEnvironmental Engineering, 116, 698-7 10.Henze, M. and Harremoes, (1983). Anaerobic treatment of wastewaters in fixed film reactors - a literaturereview. Water Science Technology, 15, 1-101.222Hilton, B.L. and Oleszkiewicz, J.A., (1987a). Toxicity of sulphides to the anaerobic treatment process.Proceedings ofthe 42nd Industrial Waste Conference, Purdue University.Hilton, B.L. and Oleszkiewicz, J.A., (198Th). A comparison of anaerobic reactors operating with andwithout the addition of suiphates. Water Pollution Research Journal ofCanada, 22, 444-45 5.Hilton, B.L. and Oleszkiewicz, J.A., (1988). Suffide-induced inhibition of anaerobic digestion. Journal ofEnvironmental Engineering, 114, 1377-1391.Hilton, M.G. and Archer, D.B., (1988). Anaerobic digestion of a sulfate-rich molasses wastewater:inhibition of hydrogen sulfide production. Biotechnology and Bioengineering, 31, 885-888.Hoeks, F.W.J.M.M., Ten Hoopen, H.J.G., Roels, H.A. and Kuenen, J.G., (1984). Anaerobic treatment of‘acid water’ (methane production in a sulfate-rich enviromnent). In: E.H. Houwink and R.R van der Meer(editors), Innovations in Biotechnology, Elsevier Science Publishers B.V., Amsterdam, 113-119.Huitman, B. and Hillstrom, R., (1986). Domsjo sulphite mill — environmental protection by thedevelopment and use of new technology., Proceedings EUCEPA Symposium, Helsinki, 19-22 May, 78-85.Imhoff, S.D. and Pfenning, N., (1983). Isolation and characterization of a mcotinic acid-degradingsulfate-reducing bacterium, Desulfococcus niacini. Archives ofMicrobialogy, 136, 194-198.Ingvorsen, K., Zehnder, A.J.B. and Jorgensen, B.B., (1984). Kinetics of sulfate and acetate uptake byDesulfobacter postgatei. Applied and Environmental Microbiology, 47, 403408.Isa, Z., Grusenmeyer, S. and Verstraete, W., (1986). Sulfate reduction relative to methane production inhigh-rate anaerobic digestion: technical aspects. Applied and Environmental Microbiology, 51, 572-579.Jackson, M., (1989). Chemi-thermomechanical Pulp (CTMP). University of British Columbia Pulp andPaper Centre lecture notes, November 20.Jarvinen, R., Vahtila, M., Mannstrom, B. and Sundholm, J., (1980). Reduced environmental load in TMP.Pulp & Paper Canada, 81, T67-T71.Jech, D.D. and Bratigam, l.A., (1983). Bioconversion of industrial residues to methane gas. In: D.L. Wise(editor), Fuel Gas Developments, CRC Press Inc., Boca Raton, Florida, 133-159.Jensen, T.L., Schleicher, 0. and Sarner, E., (1988). The Junckers industrier mill -- environmentalprotection by the development and use of new technology. TAPPI Environmental Conference, 375-3 82.Jeris, J.S. and McCarty, P.L., (1965). The biochemistry of methane fermentation using‘4C-tracers.Journal of Water Pollution Control Federation, 37,178.Jewell, W.J., (1978). Anaerobic fermentation of agricultural residue, potential for improvement andimplementation. Final Report EY-76-5-02-2981-7, Cornell University, Ithaca, New York.Jovanovic, M., Murphy, K.L. and Hall, E.R., (1986). Parallel evaluation of high rate anaerobic treatmentprocesses: retention time and concentration effects. Proceedings EWPCA Conference on WastewaterTreatment, September 15-19, Amsterdam.223Jopson, R.N., Webb, L.J. and Board, N.P., (1986). Effect of paper making additives on anaerobic effluenttreatment. PIRA Conference on Cost Effective Treatment of Papermill Effluents Using AnaerobicTechnologies, 14-15 January, Leatherhead, England, 1-11.Kaspar, H.F. and Wulinnann, K., (1978). Kinetic parameters and relative turnover of some importantcatabolic reactions in digesting sludge. Applied Environmental Microbiology, 131, 278-282.Kennedy, K.J., Andras, E., Elliot, C.M. and Methven, B., (1991a). Effect of intermittent feeding onanaerobic CTMP wastewater treatment. Environmental Technology Letters, 12, 29 1-296.Kennedy, K.J., Andras, E., Elliot, C.M. and Methven, B., (1991b). Effect of DTPA on anaerobicwastewater treatment. Canadian Journal ofCivil Engineering, 18, 53-57.Kennedy, K.J., Hamoda, M.F. and Guiot, S.G., (1988). Anaerobic treatment of leachate using fixed-filmand sludge bed systems. Journal Water Pollution Control Federation, 60, 1675-1683.Kennedy, K.J. and van den Berg, L., (1988). Stability and performance of anaerobic fixed film reactorsduring hydraulic overloading at 10-35 °C. Water Research, 16, 1391-1398.Khan, A.W., Trottier, T.M., (1978). Effect of sulfur-containing compounds on anaerobic degradation ofcellulose to methane by mixed cultures obtained from sewage sludge. Applied and EnvironmentalMicrobiology, 35, 1027-1034.Kisaalita, W.S., Pinder, K.L. and Lo, K.V., (1987). Acidogenic fermentation of lactose. Biotechnologyand Bioengineering, 30, 88-9 5.Koch, A.L., (1981). Growth measurement. In: Manual Methods for General Bacteriology, Gerhardt, P.(editor), American Society for Microbiology, Washington, 179-207.Koster, I.W., (1987). Abatement of long-chain fatty acid inhibition of methanogenesis by calciumaddition. Biological Wastes, 22, 295-301.Kristjansson, J.K., Schonheit, P., Thauer, RK., (1982). Different K values for hydrogen of methanogenicbacteria and sulfate-reducing bacteria: an explanation for the apparent inhibition of methanogenesis bysulfate. Archives ofMicrobiology, 131, 278-282.Kroiss, H. and Plahl-Wabnegg, F., (1983). Sulfide toxicity with anaerobic waste water treatment.Proceedings ofAnaerobic Waste Water Treatment Symposium held in Noordwijkerhout, The Netherlands,23 - 25 November, 72-75.Kutney, G.W., (1985). Hydrogen peroxide: stabilization of bleaching liquors. Pulp & Paper Canada, 12,182-189.Laanbroek, H.J., Geerligs, H.J., Sijtsma, L. and Veldkamp, H., (1984). Applied and EnvironmentalMicrobiology, 47, 329-334.Laanbroek, H.J. and Pfennig, N., (1981). Oxidation of short-chain fatty acids by sulfate-reducing bacteriain freshwater and in marine sediments. Archives ofMicrobiology, 128, 330-335.Labat, M. and Garcia, J.L., (1986). Study on the development of methanogenic microfiora duringanaerobic digestion of sugar beet pulp. Applied Microbiology and Biotechnology, 25, 163-168.224Lawrence, A.W. and McCarty, P.L., (1965). The role of sulfide in preventing heavy metal toxicity inanaerobic treatment. Journal Water Pollution Control Federation,. 37, 392-406.Lawrence, A.W., McCarty, P.L. and Guerin, F.J.A., (1966). The effects of sulfide on anaerobic treatment.Air and Water Pollution International Journal, 10, 207-22 1.Leach, J.M. and Thakore, A.N., (1976). Toxic constituents in mechanical pulping effluents. TAPPIJournal, 59, 129-132.Leach, J.M. and Thakore, A.N., (1977). Compounds toxic to fish in pulp mill waste streams, WaterScience Technology, 9, 787-798.Leban, M., Edwards, V.H. and Wilke, C.R, (1966). Sulfate reduction by bacteria. Journal ofFermentation Technology, 44, 334-343.Lee, J.W.Jr., Peterson, D.L. and Stickney, A.R., (1989). Anaerobic treatment of pulp and paper millwastewaters. TAPPI Environmental Conference, 473-496.LehtomAki, M. (1988). The Taman process: experiences of full-scale operation at Anjala paper mill.Water Science Technology, 20, 99-105.Lepistö, R. and Rintala, J., (1993). The removal of chlorinated phenolic compounds from chlorinebleaching effluents using thermophilic anaerobic processes. Proceedings 4th IA WQ Symposium on ForestIndustry Wastewaters, June 8-11, Tampere, Finland.Lettinga, G., (1979). Direct anaerobic treatment handles wastes effectively, Industrial Wastes, 25, 18.Lettinga, G., (1986). Future prospects for anaerobic treatment at papermills. Proceedings of the PIRAconference on Cost Effective Treaftnent of Papermill Effluents Using Anaerobic Technologies, January14-15, Leatherhead, England.Lettinga, G., Field, J.A., Sierra-Alvarez, van Lier, J.B. and Rintala, J., (1991). Future Perspectives for theanaerobic treatment of forest industry wastewaters. Water Science Technology, 3/4, 9 1-102.Lettinga, (1 and Hulshoff Pol, L., (1986). Advanced reactor design, operation and economy. WaterScience Technology, 18, 99-108.Lettinga, G., van der Geest, A. Th., Hobma, S. and Laan, J.V.D., (1979). Anaerobic treatment ofmethanolic wastes., Water Research, 13, 725-73 7.Li, A., Sutton, P.M., Corrado, J.J. and Kothari, D., (1984). Optimization of two-phase anaerobic fluidizedbed process. Proceedings: Second International Symposium on Fixed-Film Biological Processes, July 10 -12, Arlington, VA., II, 1747-1759.Lin, C.Y., Sato, K., Noike, T. and Matsumoto, J., (1986). Methanogenic digestion using mixed substrateof acetic, propiomc and butync acids. Water Research, 20, 385-394.Lo, K.V. and Liao, P., (1988). Laboratory scale studies on the mesophilic anaerobic digestion of cheesewhey in different digester configurations. Journal ofAgricultural Engineering Research, 39, 1-7.Lo, S.-N., Liu, H.W., Rousseau, S. and Lavallee, H.C., (1991). Characterization of pollutants at sourceand biological treatment of a CTIvIP effluent. Appita, 44, 133-138.225Love, L.S., (1987). Anaerobic treatment of high sulphur effluents in a unique fixed film-suspended growthreactor. 73rdAnnual Meeting of the CPPA, January 29-30, Montreal, B21-B25.Lovley, D.R, Dwayer, D.F. and Kiug, M.J., (1982). Kinetic analysis of competition between sulfatereducers and methanogens for hydrogen in sediments. Applied and Environmental Microbiology, 43,1373-1379.Lovley, D.R. and Kiug, M.J., (1983). Sulfate reducers can out compete methanogens at freshwatersulphate concentrations. Applied and Environmental Microbiology, 45, 187-192.Maat, D.Z. and Habets, L.H.A., (1986). Anaerobic treatment: a low cost and reliable wastewater treatmentprocess. 72nd CPPA Annual Meeting, B 159-B 163.Maat, D.Z. and Habets, L.H.A., (1987). Anaerobic treatment of pulp and paper effluents. 73rd AnnualMeeting ofthe CPPA, B295-B299.Maatta, R.K., (1985). Anaerobic wastewater treatment processes. Water Science Technology, 17, 53-59.Mackie, R.L. and Bryant, M.P., (1981). Metabolic activity of fatty acid-oxidizing bacteria and thecontribution of acetate, butyrate, and CO2 to methanogenesis in cattle waste at 40 and 60 °C. AppliedEnvironmental Microbiology, 41, 1363-1373.MacLean, B., de Vegt, A. and van Driel, E., (1990). Full-scale anaerobic/aerobic treatment ofTMPIBCTMP effluent at Quesnel River Pulp. TAPPI Environmental Conference, 647-661.Mah, RA., (1983). Interactions of methanogens and non-methanogens in microbial ecosystems.Proceedings ofthe Third International Symposium on Anaerobic Digestion, August 14-19, Boston Mass.,13-22.Mah, RA., (1982). Methanogenesis and methanogenic partnerships. PhiL Trans. Royal Society ofLondon, B 297, 599-616.Malina, IF., (1964). Thermal effects on completely mixed anaerobic digestion. Water & Sewage Works,11, 52.Maillacheruvu, K.Y., Parkin, G.F., Ye Peng, C., Kuo, W-C., Oonge, Z.I. and Lebduschka, V., (1993).Suffide toxicity in anaerobic systems fed sulfate and various organics. Water Environment Research, 65,100-109.Malinen, R., Virkola, N.-E., and Turunen, B., (1985), Effluent characterization and treatment ofchemithermomechanical pulping. Proceedings of the SPCI International Mechanical PulpingConference, 267-273.Maly, IF. and Fadrus, H., (1971). Influence of temperature on anaerobic digestion. Journal of the WaterPollution Control Federation, 43, 64 1-650.Martens, C.S. and Berner, RA., (1974). Methane production in the interstitial waters of sulfate-depletedmarine sediments. Science, 185, 1167-1169.Maiee, J.P., Strydom, W.F., (1983). Biological sulphate removal in an upflow packed reactor. WaterResources, 19, 1101-1106.226Massey, ML. and Pohiand, F.G., (1978). Phase separation of anaerobic stabilization by kinetic controls.Journal ofthe Water Pollution Control Federation, 50, 2204-2222.McBride, B.C. and Wolfe, R.S., (1971). A new coenzyme of methyl-transfer, coenzyme M. Biochemistry,10, 2137.McCarty, P.L., (1964). Anaerobic waste treatment fundamentals. Public Works, 95, 9 1-94.McCarthy, P.3., Kennedy, K.J. and Droeste, R.L., (1990). Role of resin acids in the anaerobic toxicity ofchemithermomechanical pulp wastewater. Water Research.McCartney, D.M. and Oleszkiewicz, J.A., (1992). Sulphide inhibition of anaerobic degradation of lactateand acetate, Water Pollution Research Journal ofCanada.McCubbin Consultants Inc., (1992). Best Available Technology for the Ontario Puly and Paper Industry.Water Resources Branch, Ontario Ministry of the Environment.McFarland, M.J. and Jewell, W.J., (1990). The effect of sulfate reduction on the therinophilic (55 °C)methane fermentation process. Journal ofIndustrial Microbiology, 5, 247-25 8.Mclnerney, M.J. and Bryant, M.P., (1981). Review of methane fermentation fundamentals. In: D. Wise(editor). Fuel Gas Production From Biomass. Volume I. CRC Press, Boca Raton, Florida, 19-46.McKinney, RE., (1962). Microbiology for Sanitary Engineers, McGraw-Hill, New York.McKinney, R.E., (1984). Anaerobic treatment concepts. TAPPI Environmental Conference, 163-172.Mehrotra, S., Springer, AM., Eralp, A.E. and Kreissl, J.P., (1987). The effect of various sulphur specieson the treatment efficiency of anaerobic processes. TAPPI Environmental Conference, 155-163.Middileton, A.C. and Lawrence, A.W., (1977). Kinetics of microbial sulfate reduction. Journal WaterPollution Control Federation, 49, 1659-1670.Minami, K., Tanimoto, Y., Tasaki, M., Ogawa, S., Okamura, K., (1988). Influence of sulfate on methanefermentation in a defined medium. Water Science Technology, 20, 451-453.Minami, K., Tanimoto, Y., Tasaki, M., Ogawa, S., Okamura, K., (1988). Influence of pH on methane andsuffide production from methanol. Journal ofFermentation Technology, 66, 117-121.Mitchell, R, (1974). Introduction to Environmental Microbiology, Prentice-Hall, Englewood Cliffs, N.J.Morris, J.G., (1976). Oxygen and the obligate anaerobe. Journal ofApplied Bacteriology, 40, 229-244.Mosey, F.E., Swanwick, J.D. and Hughes, D.A., (1971). Factors affecting the availability of heavy metalsto inhibit anaerobic digestion. Journal Water Pollution Control Federation, 668-680.Mounifort, D.O. and Asher, RA., (1979). Effect of inorganic suffide on the growth and metabolism ofMethanosarcina barberi strain DM., Applied and Environmental Microbiology, 37, 670-675.Mudrack, V.K. and Kunst, S., (1982). Untersuchungen zum zweistuflgen betrieb von reaktoren zuranaeroben behandlung von kohienhydraten. Z. Wasser Abwasser Forsch, 15, 277-287.227Mueller, J.C., Leach, J.M. and Walden, C.C., (1976). Biological detoxification of pulp and paper milleffluents. Abstracts ofpapers, Fifth internationalfermentation symposium, Berlin, 425.Mulder, A., (1984). The effects of high sulphate concentrations on the methane fermentation ofwastewater. In: E.H. Houwink and R. van der Meer (Editors) Innovations in Biotechnology, ElsevierScience Publishers, Amsterdam, 133-143.Nakatsukasa, W. and Akagi, J.M., (1969). Thiosulfate reductase isolated from Desufotomaculumnigrficans. Journal ofBacteriology, 98, 429.Nanninga, H.J. and Gottschal, J.C., (1987). Properties of Desulfovibrio carbinolicus sp. nov. and othersulfate-reducing bacteria isolated from an anaerobic purification plant. Applied and EnvironmentalMicrobiology, 53, 802-809.Neal, A.L., Weinstock, J.O. and Lampen, JO., (1965). Mechanism of fatty acid toxicity for yeast. JournalofBacteriology, 90, 126.Newport, P.J. and Nedwell, D.B., (1988). The mechanism of inhibition of Desulfovibrio andDesufotomaculum species by selenate and molybdate. Journal ofApplied Bacteriology, 65, 419-423.Norqvist, A. and Roffey, R., (1983). Alternative method for monitoring the effect of inhibitors on sulfatereduction. Journal ofGeneralAppliedMicrobiology, 29, 334-344.Novaes, RV.F., (1986), Microbiology of anaerobic digestion. Water Science Technology, 18, 1-14.O’Connor, B.I., Kovacs, T.G. and Voss, R.H., (1992). The effect of wood species composition on thetoxicity of simulated mechanical pulping effluents. Environmental Toxicology and Chemistry, 11, 125 9-1270.Oleszkiewicz, J.A. and Hilton, B.H., (1986). Anaerobic treatment of high-sulfate wastes. CanadianJournal ofCivil Engineering, 13, 4323-428,Oleszkiewicz, J.A. and Koziarski, S., (1982). Anaerobic wastewater treatment in flow-through digesterswithout sludge recycle. Agricultural Wastes, 365-3 86.Oleszkiewicz, J.A. and Rornanek, A., (1989). Granulation in anaerobic sludge bed reactors treating foodindustry wastes. Biological Wastes, 27, 217-235.Olthof M., Kelly, W.R., Oleszkiewicz, J.A. and Weinreb, H.G., (1985). Development of anaerobictreatment process for wastewaters containing high sulfates. Proceedings of the 40th Industrial WasteConference, Purdue University, 871-877.Oreniland, R.S. and Taylor, B.F., (1978). Sulfate reduction and methanogenesis in marine sediments.Geochimica Cosmochimica Acta, 42, 209-214.Orivuori, 3., (1985). Forest industry anaerobic wastewater treatment and its future development. WaterScience Technology, 17, 265-270.Parkin, G.F., Lynch, N.L., Kuo, W.C., van Kueren, E.L. and Bhattacharya, S.K., (1990). Interactionbetween sulfate reducers and methanogens fed acetate and propionate. Research Journal of the WaterPollution Control Federation, 62, 780.228Parkin, G.F. and Speece, RE., (1984). Anaerobic biological waste treatment. Chemical EngineeringProgress, December, 55-58.Patel, G.B., (1984). Characterization and nutritional properties of Methanothrix concilii Sp. Nov., amesophilic, acetoclastic methanogen. Canadian Journal ofMicrobiology, 30, 1383-1396.Pearson, J., (1985). IMPC spotlights new technology. Pulp and Paper, October.Peck, H.D. Jr. and LeGall, J., (1982). Biochemistry of dissimilatory sulphate reduction, PhilosophicalTransactions ofthe Royal Society i,fLondon, B298, 443-466.Pfenning, N., Widdel, F. and Truper, 11G., (1981). The dissimilatory sulphate reducing bacteria. In: M.P.Starr, H. Stoip, H.G. Truper, A. Balows, and H.G. Schegel (editors), The procarvotes. Springer-Verlag,Berlin.Phelps, T.J., Conrad, R. and Zeilcus, J.G., (1985). Sulfate-dependent interspecies H2 transfer betweenMethanosarcina barkeri and Desulfovibrio vulgaris during coculture metabolism of acetate or methanol.Applied and Environmental Microbiology, 50, 589-594.Pichon, M., de Choudens, C., Meyer, F. and Francois, E., (1986). Anaerobic treatment of CTMPwastewater. Proceedings EUCEPA Symposium, Helsinki, 19-22 May, 131-139.Pichon, M., Najean, H. and Rouger, J., (1987). Anaerobic treatment of softwood suffite CTh4Pwastewater. Proceedings Fourth International Symposium on Wood and Pulping Chemistry, April 27-30,Paris, 1, 1-4.Pichon, M., Rouger, I. and Junet, E., (1988). Anaerobic treatment of sulphur-containing effluents. WaterScience Technology, 20, 133-141.Pipyn, P., Eeckhaut, M., Ombregt, J.P., Teraoka, H., Kishimoto, Y., Sakamoto, H., Kuroda, I., andMasue, Y., (1985). Anaerobic treatment of kraft pulp mill condensate. TAPPI Environmental Conference,173-178.Pipyn, P. and Verstraete, W., (1981). Lactate and ethanol as intermediates in two-phase anaerobicdigestion, Biotechnology and Bioengineering, 23, 1145-1154.Pohland, F.G. and Bloodgood, D.E., (1963). Laboratory studies on mesophilic and thermophilic anaerobicsludge digestion. Journal Water Pollution Control Federation, 35, 11.Postgate, J.R., (1979). The Sulfate-Reducing Bacteria. Cambridge University Press, Cambridge.Postgate, J.R., (1984). The Sulfate-Reducing Bacteria. 2nd edition, Cambridge University Press,Cambridge.Potapenko, S.A., Bochagova, O.P., Voloshina, L.A., Starodub, I.M., Matjusha, GA. and Cheblina, G.P.,(1993). On some features of CTMP wastewater anaerobic treatment, Proceedings 4th IA WQ Symposiumon Forest Industry Wastewaters, June 8-11, Tampere, Finland.Pretorius, W.A., (1987). A conceptual basis for microbial selection in biological wastewater treatment.Water Research, 8, 891-894.Price, E.C., (1985). The microbiology of anaerobic digestion. In: P.N. Cheremisinoff and R.P. Ouellette(editors), Biotechnology Applications and Research, Technomic Publishing Co., Lancaster, Pennsylvania.229Prong, C,F. and Chmelauskas, A.J., (1988). Demonstration of UASB anaerobic treatment technology forNSSC wastewater treatment at Sturgeon Falls, Ontario. Preprints, Treatment of ChemithermomechanicalPulping Effluents, October 27, Vancouver, B.C., 79-91.Puar, R., (1991). Personal communication at Quesnel River Pulp.Puhakka, J., (1990). Amenability of Pulp Mill Wastes to Anaerobic Treatment. Academic dissertation inGeneral Microbiology, Water and Environmental Engineering, Tampere University of Technology,Finland.Ptihakka, J.A., Ferguson, J.F., Benjamin, M.M. and Salkinoja-Salonen, M.,(1989). Sulfur reduction andinhibition in anaerobic treatment of simulated pulp mill wastewater. Systems in Applied Microbiology, 11,202-206.Puhakka, J.A., Rintala, J.A., Vuoriranta, P., (1985). Influence of sulfur compounds on biogas productionfrom forest industry wastewater. Water Science Technology. 17, 28 1-288.Puhakka, J.A., Salkinoja-Salonen, M., Ferguson, J.F. and Benjamin, M.M., (1990). Effect of molybdateions on methanation of simulated and natural waste-waters. Applied Microbiology and Biotechnology, 32,494498.Puhakka, J.A., Salkinoja-Salonen, M., Ferguson, J.F. and Benjamin, M.M., (1990). Carbon flow inacetotrophic enrichment cultures from pulp mill effluent treatment. Water Research, 11.Puhakka, J.A., Viitasaari, M.A., Latola, P.K. and Maatta, R.K., (1988). Effect of temperature onanaerobic digestion of pulp and paper industry wastewater sludges. Water Science Technology, 20, 193-201.Rankin, A., MacLean, B., Irvine, GA. and Stevens, G., (1992). Pilot studies on activated sludge atQuesnel River Pulp. Proceedings of the Spring Conference of the Canadian Pulp and Paper Association,May 13-16, Jasper, Alberta.Rantala, P. and Vaananen, P., (1985). Cost comparison of aerobic and anaerobic, wastewater treatmentsystems. Water Science Technology, 17, 255 - 264.Rekunen, S., (1985). Anaerobic wastewater treatment system operates at very low cost. Pulp & Paper, 59,122-125.Richardson, D.A., Andras, E. and Kennedy, K.J., (1990). Anaerobic toxicity of fines in chemithermomechanical pulp wastewaters: a batch assay - reactor study comparison. Presented at the 3rdIA WPRC Symposium on Forest Industry Wastewaters, Tampere, Finland, June 5 - 8.Richardson, D.E. and Bloom, H., (1983). Chemical composition of treated thermomechanical pulpeffluent from an activated sludge plant. Appita, 36, 456-460.Rintala, J.A. and Lepisto, S.L., (1992). Anaerobic treatment of thermomechanical pulping whitewater at35 -70 °C, Water Research, 26, 1297-1305.Rintala, J., Samz Martin, J.L. and Lettinga, G., (1991). Thermophilic anaerobic treatment of sulfate-richpulp and paper integrate process water. Water Science Technology, 3/4. 149-160.230Rintala, J. and Vuoriranta, P., (1988). Anaerobic-aerobic treatment of thermomechanical pulpingeffluents. TAPPI Journal, 9, 20 1-206.Rintala, 3., Sanz Martin, J.L. and Lettinga, G., (1991). Thermophilic anaerobic treatment of sulfate-richpulp and paper integrate process water. Water Science Technology, 24, 149-160.Rinzema, A. and Lettinga, G., (1988). The effect of sulphide on the anaerobic degradation of propionate.Environmental Technology Letters, 9, 83-88.Rogers, I., Mahood, H., Servizi, J. and Gordon, R., (1979). IdentiI’ing extractives toxic to aquatic life,Pulp and Paper Canada, T286-T290.Ronnow, PH., Gunnarsson, L.A.H., (1981). Suffide-dependent methane production and growth of athermophilic methanogemc bacterium. Applied and Environmental Microbiology, 42, 580-584.Rosen, B. and Gunnarsson, L., (1986). Treatment of mill wastewaters with anaerobic bed filters.Proceedings ofthe PIRA conference on Cost Effective Treatment ofPapermill Effluents Using AnaerobicTechnologies, January 14-15, Leatherhead, England.Saleh, A.M., MacPherson, R. and Miller, J.D.A., (1964). The effect of inhibitors on sulphate reducingbacteria: a compilation. Journal ofApplied Bacteriology, 27, 281-293.Salkinoja-Salonen, M.S., Hakulinen, R, Silakoski, L., Valo, L. and Apajalahti, J., (1983). Treatment ofpulp and paper industiy wastewaters in an anaerobic reactor: from bench size units to full scale operation.Proceedings, 3rd International Symposium on Anaerobic Digestion, 14-19 August, Boston, 107-121.Samer, E., Hultman, B.G. and Bergiund, A.E., (1988). Anaerobic treatment using treatment forcontrolling H2S toxicity. TAPPI Journal, 41-45.Scherer, P. and Sahm, H., (1981). Influence of sulphur-containing compounds on the growth ofMethanosarcina barkeri in a defined medium. European Journal of Applied Microbiology and -Biotechnology, 12, 28-35.Schneider, A., Gerbi, V. and Redoglia, M., (1987). A rapid HPLC method for separation anddetermination of major organic acids in grape musts and wines. American Journal of Enology andViticulture, 38, 15 1-155.Schneider, E., Wilson, R.W. and Frenette, E.G., (1986). Anaerobic pilot plant experience at ConsolidatedBathurst Inc., Bathurst, N.B. mill. Proceedings of 73rd CPPA Meeting, B43-B49.Schiegel, H.G., (1969). Allgemeine Microbiologie 4, Auflage.Schnell, A., Donca, J., Ho, C., Ashikawa, M., Munnoch, G. and Hall, E.R., (1990). Anaerobic andaerobic pilot-scale effluent detoxification studies at an integrated newsprint mill. Pulp & Paper Canada,91, 75-80.Schonheit, P., Kristjansson, J.K., Yhauer, R.K., (1982). Kinetic mechanisms for the ability of sulfatereducers to out-compete methanogens for acetate. Archives ofMicrobiology, 132, 285-288.Schraa, G. and Jewell, W.J., (1984). High rate conversions of soluble organics with a thermophilicanaerobic attached film expanded bed. Journal Water Pollution Control Federation, 56, 226-232.231Servizi, l.A. and Gordon, R.W., (1986). Detoxification of TMP and CTMP effluents alternating in a pilotscale aerated lagoon. Pulp & Paper Canada, 11, 42-46.Shere, S.M. and Daly, P.G., (1982). High rate biological treatment of TMP effluent. Pulp and PaperCanada, 83, T91-T95.Sierra-Alvarez, R., Field, J.A., Kortekaas, S. and Lettinga, G., (1993). Overview of the anaerobic toxicitycaused by organic forest industry wastewater pollutants, Proceedings 4th IA WQ Symposium on ForestIndustry Wastewaters, June 8 - 11, Tampere, Finland.Sierra-Alvarez, R, Kortekaas, S., van Eckert, M. and Lettinga, G., (1990). The anaerobic biodegradabilityand methanogenic toxicity of pulping wastewaters, Proceedings 3rd IA WQ Symposium on Forest IndustryWastewaters, Tampere, Finland.Smart, J. and Boyko, B.I., (1977). Full scale studies on the thermophilic anaerobic digestion process.Environment Canada Research Report No. 59, Ottawa, Ontario.Smith, M.R.S., Zinder, S.H. and Mah, R.A., (1980). Microbial methanogenesis from acetate. ProcessBiochemistry, 15, 34-39.Smith, P.H. and Mali, R.A., (1966). Kinetics of acetate metabolism during sludge digestion. Applied andEnvironmental Microbiology, 14, 368-371.Smith, R.L. and Kiug, M.J., (1981). Electron donors utilized by sulfate-reducing bacteria in eutrophic lakesediments. Applied Environmental Microbiology, 42, 116-121.Sorensen, J., Christensen, D. and Jorgensen, B.B., (1981). Volatile fatty acids and hydrogen as substratesfor sulfate-reducing bacteria in anaerobic marine sediments. Applied Environmental Microbiology, 42, 5-11.Speece, R.E., (1983). Anaerobic biotechnology for industrial wastewater treatment. Environmental.Science and Technology, 17, 416A-427A.Springer, A.M., (1986). Industrial Environmental Control Pulp and Paper Industry, John Wiley & Sons,Chapters 2, 3 and 6.Stander, G.J. and Elsworth, J.F., (1950). Effluents from fermentation Industries. Part III -- the influenceof high concentrations of suiphates on the anaerobic digestion method of treatment at thermophilic (55°C) and mesophilic temperatures. Journal ofthe Institute ofSewage Purification, Part 3, 303.Swan, E.P., (1973). Resin acids and fatty acids of Canadian pulpwoods -- a review of the literature.Department of the Environment, Canadian Forestry Service, Information Report VP-X-1 15.Tammoto, Y., Tasaki, M., Okamura, K., Yamaguchi, M., Minami, K., (1989). Screening growthinhibitors on sulfate-reducing bacteria and their effects on methane fermentation. Journal ofFermentationand Bioengineering, 68, 353-3 59.Taylor, B.F. and Oremland, R.S., (1987). Depletion of adenosine triphosphate in Desulfovibrio byoxyanions of group VI elements. Current Microbiology, 3, 10 1-103.Thakore, A., Collins, G. and Smiley, D.W., (1977). Toxicity and its behaviour on pulp fibres. CPPASpring Conference, Whistler, B.C., May 14-16, 1-4.232Toerien, D.F. and Hattingh, W.H.J., (1969). Anaerobic digestion I. The microbiology of anaerobicdigestion., Water Research, 3, 385-416.Trudinger, P.A., (1981). Assimilatory and dissimilatory metabolism of inorganic sulphur compounds bymicro-organisms. Advances in Microbial Physiology, 111-158.Turk, S., (1988). Options for treatment of CTMP effluents. Paper presented at Seminar on Treatment ofCTMP Effluents, 27 October, Vancouver, 15-46.Ueki, A., Matsuda, K. and Ohtsuki, C., (1986). Sulfate-reduction in the -anaerobic digestion- of animal -waste. Journal ofGeneral Applied Microbiology, 32, 111-123.Urbantas, R.G. and MacEwen, H.D., (1985). Characterization of effluents from chemimechanical,chemithermomechanical and thermomecharncal pulping of black spruce. preprints, 71st Annual Meeting,Technical Section, Canadian Pulp and Paper Association, A213-A2 19.van den Berg, L., Kennedy, K.J. and Samson, (1985). Anaerobic downflow stationary fixed film reactors:performance under steady-state and non-steady state conditions. Water Science Technology, 17, 89-102.van den Berg, L., Lamb, K.A., Murray, W.D., Annstrong, D.W., (1980). Effects of sulphate, iron andhydrogen on the microbiological conversion of acetic acid to methane. Journal ofApplied Bacteriology.48, 437-447.van den Berg, L., Patel, G.B., Clark, D.S. and Lentz, C.P., (1976). Factors affecting rate of methaneformation from acetic acid by enriched methanogenic cultures. Canadian Journal of Microbiology, 22,1312-1319.van Velsen, A.F.M., Lettinga, G. and Ottelander, (1979). Anaerobic digestion of piggery waste. 3.Influence of temperature. Netherlands Journal ofAgricultural Sciences, 27, 255-267.Varel, V.H., Isaacson, H.R. and Bryant, M.P., (1977). Thermophilic methane production from cafflewaste. Applied Environmental Microbiology, 33, 2198.Veldkamp, H. and Jannasch, H.W., (1972). Mixed culture studies with the chemostat. Journal ofAppliedChemistry and Biotechnology, 22, 105-123.Virkola, N.E. and Honkanen, K., (1985). Wastewater characteristics. Water Science Technology, 17, 1-28.Webb, L.J., (1983). Anaerobic biological treatment of wastewaters from the pulp and paper industry.Proceedings, PIRA Jnternational Conference on Biotechnology in the Pulp and Paper Industry,September 12-14, London, 1-29.Webb, L.J., (1985). Characteristics of paper/board mill wastewaters relevant to anaerobic treatment. WaterScience Technology, 17, 29-38.Welander, T., (1988). An anaerobic process for treatment of CTMP effluent. Water Science Technology,20, 143-147.Welander, T., (1989). Anaerobic Treatment of CTMF Effluent. Doctoral Dissertation, Department ofApplied Microbiology, Lund University, Sweden.233Welander, T. and Andersson, P.-E., (1985). Anaerobic treatment of wastewater from the production ofchemi-thermomechanical pulp. Water Science Technology, 17, 103-111.Wellinger, A. and Wuhrmann, K., (1977). Influence of suffide compounds on the metabolism ofMethanobacterium strain AZ. Archives ofMicrobiology, 115, 13-17.Widdel, F., (1988). Microbiology and ecology of sulfate- and sulfur-reducing bacteria. In: Biology ofAnaerobic Microorganisms, A.J.B. Zehnder (editor), 1, 467-585, John Wiley & Sons, New York, USA.Widdel, F. and Pfenrnng, N., (1977). A new anaerobic, sporing, acetate -oxidizing, -sulphate- reducing- - - -bacterium, Desulfotomaculum (emend.) acetoxidans. Archives ofMicrobiology, 112, 119-122.Widdel, F. and Pfenrnng, N., (1981). Studies on dissimilatory sulfate-reducing bacteria that decomposefatty acids I. Isolation of new sulfate-reducing bacteria enriched with acetate from saline environments.Description ofDesuifobacterpostgatei gen. nov., sp. nov. Archives ofMicrobiology, 129, 395-400.Wiegant, W.M., Claassen, J.A., Borghans, A.J.M.L. and Lettinga, G., (1986). High rate thermophilicanaerobic digestion for the generation of methane from organic wastes. Water Science Technology, 392-409.Wildenauer, F.X. and Winter, J., (1985). Anaerobic digestion of high-strength acidic whey in a pH-controlled up-flow fixed film loop reactor. Applied Microbiology and Biotechnology, 22, 367-372.Wilhelm, E., Battino, R. and Wilcock, J. (1977). Low-pressure solubility of gases in liquid water.Chemistry Reviews, 77, 2 19-262.Wilkins, A.L. and Panadam, S., (1987). Extractable organic substances from the discharges of a NewZealand pulp and paper mill, Appita, 40, 208-212.Wilkinson, S.G., (1967). The sensitivity of pseudomonads to ethylenediaminetraacetic acid. Journal ofGeneral Microbiology, 47, 67-76.Wilson, L.E., (1981). Digestion design concepts. In: D.L. Wise (editor), Fuel Gas Developments, CRCPress Inc., Boca Raton, Florida, 63-109.Wilson, R.W. and Frenette, E.G., (1988). Development of anaerobic/aerobic treatment at ConsolidatedBathurst Inc., Proceedings ofthe Seminar on Treatment ofCTMP Effluents, October 27, 93-110.Wilson, R.W., Murphy, K.L. and Frenette, E.G., (1987). Aerobic and anaerobic treatment of NSSC andCTMP effluent. Pulp and Paper Canada, 88, T4-T8.Winfrey, M.R. and Ward, D.M., (1983). Substrates for sulfate reduction and methane production in intertidal sediments. Applied and Environmental Microbiology, 45, 193-199.Winfrey, M.R. and Zeikus, J.G., (1977). Effect of sulfate on carbon and electron flow during microbialmethanogenesis in freshwater sediments. Applied and Environmental Microbiology, 33, 275-281.Wu, W.-M., Hickey, R.F. and Zeikus, J.G., (1991). Characterization of metabolic performance ofmethanogenic granules treating brewery wastewater: role of sulfate-reducing bacteria. Applied andEnvironmental Microbiology, 57, 3434-3449.Wuhrmann, K., (1982). Ecology of methanogemc systems in nature. Experientia, 38, 193-198.234Yamaguchi, M., Minaini, K., Tammoto, Y. and Okamura, K., (1989). Effects of volatile fatty acids onmethanogenesis of methanol and pre growth with methanol on acetate utilization by methanogens.Journal ofFermentation and Bioengineering, 68, 428-432.Yoda, M., Kitagava, M. and Miyaji, Y., (1987). Long term competition between sulfate-reducing andmethane-producing bacteria for acetate in anaerobic bioflim. Water Research, 21, 1547-15 56.Zehnder, A.J.B., (1981). Ecology of methane fonnation. In: R Mitchell (editor), Water PollutionMicrobiology, Vol. 2, Wiley, New York, 349-376.Zehnder, A.J.B., Huser, B.A., Brock, T.D. and Wuhnnann, K., (1980). Characterization of an acetatedecarboxylating, non-hydrogen oxidizing methane bacterium. Archives ofMicrobiology, 24, 1-11.Zehnder, A.J.B., Ingvorsen, K. and Marty, T., (1982). Microbiology of methane bacteria. In: AnaerobicDigestion 1981, Hughs etal. (editors), Elsevier, Amsterdam.Zehnder, A.J.B. and Wubrmann, (1977). Physiology of a Methanobacterium strain AZ. Archives ofMicrobiology, 111, 199-205.Zeikus, J.G., (1977). The biology of methanogenic bacteria. Bacteriological Reviews, 41, 514-541.Zeikus, J.G. and Wolfe, R.S., (1972). Methanobacterium thermoautotrophicum sp. n., an anaerobic,autotrophic, extreme thermophile. Journal ofBacteriology, 109, 707.Zinder, S.H., (1986). Thermophilic waste treatment systems. In: Thermopiles: General Molecular andAtplied Biology, T.D. Brock (editor). Wiley-Interscience, New York, 25 7-277.Zinder, S.H. (1988). Conversion of acetic acid to methane by thermophiles. In: Proceedings FifthInternational Symposium on Anaerobic Digestion, A. Tilche and A. Rozzi (editors), Monduzzi Editore,Bologna, Italy.Zinder, S.H. and Mali, R.A., (1979). Isolation and characterization of a thermophilic strain ofMethanosarcina unable to use H2-C0 for methanogenesis, Applied Environmental Microbiology, 38,996-1008.Zoetemeyer, RJ., Arnoldy, P., Cohen, A. and Boelhouwer, C., (1982). Influence of temperature on theanaerobic acidification of glucose in a mixed culture forming part of a two-stage process. Water Research,16, 3 13-321.235Appendix 1: Carbon BalanceCarbon enters the system in the feedstock and via the sodium carbonate pHadjusting solution. Carbon exits the system in the liquid effluent, in the form of methaneand carbon dioxide evolved in the gas as well as that dissolved in the liquid effluent, in theform of precipitated organic compounds and as the net growth of biomass. A carbonbalance is completed for the following reasons:1) to provide an indication of the accuracy of the total carbon, gas volume and gascomposition assays, and2) in the absence of the ability to take a representative and homogeneous sample due tothe unmixed retained biomass nature of the reactor system, the unaccounted for carbonprovides an indication of the organic compounds which were precipitated from solution aswell as the sludge yield. Wilson et al. (1987) for example commented that accuratepredictions of sludge production are notoriously difficult to achieve in small pilot systems.The following Tables include carbon balances for the three sulphur managementstrategies. Note that due to the difficulty of obtaining representative samples whichcontained solids and liquids from the fixed and suspended biomass reactor system, theTOC measured was that of centrifuged samples. Therefore, the dissolved carbonates, theorganic compounds present as biomass as well as those organic compounds which wereprecipitated from solution are not included in the carbon balance.236Table 67: Carbon Balance for Acid Phase Reactor pH Control Experiments at 35 °CHRT Total Carbon Lmg C/liAcid Phase [days] Soluble CH4(g) C02(g) Total C %Reactor pH TOC Recovered5.5 Feed 1065 - - 10650.6 1011 54 65 1130 1061.2 895 153 186 1234 1161.8 926 117 142 1185 1116.5 Feed 1050 - - 10500.6 903 41 50 994 951.2 882 103 147 1132 1071.8 724 117 142 983 947.0 Feed 2415 - - 24150.6 2077 63 77 2217 901.2 1787 158 192 2137 871.8 1860 149 182 2191 897.5 Feed 3235 - - 32350.6 2459 72 88 2619 811.2 1456 158 193 1807 561.8 1197 329 402 1928 608.0 Feed 3175 - - 31750.6 2445 50 61 2556 811.2 1873 99 121 2093 661.8 1937 140 171 2248 71237Table 68: Carbon Balance for Acid Phase Reactor pH Control Experiments at 55°CHRT Total Carbon [mg C/I]Acid Phase [days] Soluble CH4(g) C02(g) Total C %Reactor pH TOC Recovered7.0 Feed 3400 - - 34000.6 2652 4 5 2661 781.2 2822 10 13 2895 841.8 3026 9 12 3047 907.5 Feed 2930 - - 29300.6 2344 14 17 2375 811.2 2403 4 5 2412 821.8 2227 5 7 2239 768.0 Feed 3355 - - 33550.6 2717 4 5 2726 811.2 2918 6 7 2931 871.8 2683 7 9 2699 80238Table 69: Carbon Balance for Molybdate Addition Experiments at 35°CMolybdate HRT Total Carbon [mg C/I]Addition [daysi Soluble CH4(g) C02(g) Total C %[mM] TOC Recovered0 Feed 2754 - - 2754batch 1 0.6 1955 41 50 2046 751.2 2066 68 83 2217 811.8 1818 54 66 1938 850.1 Feed 2745 - - 2745batch 1 0.6 2086 50 61 2197 801.2 1784 95 116 1995 731.8 1812 117 143 2072 750.5 Feed 2735 - - 2734batch 1 0.6 1696 41 50 1787 651.2 1614 72 88 1774 651.8 1586 108 132 1826 670.75 Feed 2635 - - 2635batch 1 0.6 1765 23 28 1816 691.2 1633 36 44 1713 651.8 1607 59 72 1738 661.0 Feed 2395 - - 2395batch 1 0.6 1868 - - -1.2 1533 - - -1.8 1700 - - -1.0 Feed 2124 - - 2124batch 2 0.6 1678 14 17 1709 801.2 1657 27 33 1717 811.8 1551 45 55 1651 781.0 Feed 2021 - - 2021batch 3 0.6 1950 4 5 1959 971.2 1935 32 39 2006 991.8 1920 36 44 2000 99239Table 70: Carbon Balance for Iron Plus Molybdate Addition Experiments at 35 °CIron + HRT Total Carbon Lmg C/liMolybdate [days] Soluble CH4(g) C02(g) Total C %Addition TOC RecoveredOmMMo+ Feed 2690 - - 26905mg/i 0.6 2098 16 18 2132 80FeC13*6H20 1.2 2044 29 38 2112 79(control) 1.8 2044 58 65 2167 811.OmMMo+ Feed 2020 - - 20205mg/i 0.6 1616 5 9 1630 82FeC136H2O 1.2 1576 33 40 1649 821.8 1333 34 42 1409 701.OmMMo+ Feed 1965 - - 1965100mg/i 0.6 1454 9 11 1474 75FeC136H2O 1.2 1336 30 36 1402 711.8 1359 35 39 1433 731.OmMMo+ Feed 2575 - - 2575200mg/i 0.6 2009 12 14 2035 79FeC136H2O 1.2 1957 27 36 2020 781.8 1957 42 50 2049 80240Table 71: Carbon Balance for Molybdate Addition Experiments at 55 °CMolybdate HRT Total Carbon jmg C/i]Addition [days] Soluble CH4(g) C02(g) Total C %[mM] TOC Recovered0 Feed 2530 - - 25300.6 2024 4 6 2034 801.2 2075 5 7 2092 831.8 1948 12 17 1977 780.5 Feed 2425 - - 24250.6 1964 4 5 1973 811.2 1916 6 8 1930 801.8 2013 15 17 2045 841.0 Feed 2790 - - 27900.6 1981 6 7 1994 711.2 1981 14 15 2010 721.8 1897 18 20 1935 70241Table 72: Carbon Balance for Gas Stripping Experiments at 35°CGas HRT Total Carbon jmg C/IlStripping [days] Soluble CH4(g) C02(g) Total C %Experiment TOC Recoveredwater scrubber Feed 2990 - - 2990control 0.6 2332 14 17 2363 79batch 1 1.2 2272 32 36 2340 781.8 2272 51 64 2387 80ferric chloride Feed 2990 - - 2990scrubber 0.6 2044 11 14 2069 69batch 1 1.2 1850 75 85 2010 671.8 1395 60 66 1521 51ferric chloride Feed 3475 - - 3475scrubber 0.6 2363 9 11 2383 69batch 2 1.2 1842 58 63 1963 561.8 1840 86 102 2028 58ferric chloride Feed 3540 - - 3540scrubber 0.6 2018 14 17 2049 58batch 2 1.2 1735 78 86 1899 53+sulphate 1.8 1558 105 127 1790 51242Table 73: Carbon Balance for Gas Stripping Experiments at 55 °CGas HRT Total Carbon Lmg C/l]Stripping [days] Soluble CH4(g) C02(g) Total C %Experiment TOC Recoveredwater scrubber Feed 2940 - - 2940control 0.6 2146 5 6 2157 741.2 2117 4 5 2126 721.8 2000 15 17 2032 69ferric chloride Feed 2940 - - 2940scrubber 0.6 2186 7 9 2202 751.2 2117 10 11 2138 731.8 2220 28 30 2278 77243Appendix 2: Sulphur BalanceThe mean sulphate, sulphide, suiphite and thiosulphate concentrations aretabulated below as their sulphur equivalents. The mass of hydrogen sulphide which waspartitioned in the gas phase was calculated from the total dissolved sulphide concentrationand the pH in the methane phase reactor liquor as well as the volumetric loading rateaccording to equation 1 (page 67) and Henry’s law. The mass of sulphur which wasprecipitated as metal sulphides was not quantified but can be expected to be significant,especially for the molybdate plus iron experiments.Eis et al. (1983) reported that only 10 to 60% of the influent sulphate sulphur wasaccounted for as sulphide in anaerobic processes. These results demonstrated a similarlevel of precision. Sulphur recovery ranged between 46 and 100%. In the cases where thesamples were spoiled prior to assay of sulphite and thiosuiphate, the known values aretabulated but no sulphur recovery efficiency is calculated.244Table 74: Sulphur Balance for Acid Phase Reactor pH Experiments at 35 °CAcid Phase HRT Sulphur Compound [mg S/i]Reactor pH [days] S042 S2 H2S g S03 S203 Total* * Recovered5.5 Feed 200 2 - 2020.6 54 16 6 761.2 26 16 5 471.8 20 15 5 406.5 Feed 214 3 - 2170.6 54 24 9 871.2 54 25 7 861.8 47 19 5 717.0 Feed 356 3 - 3590.6 182 40 10 2321.2 39 87 21 1471.8 46 77 19 1427.5 Feed 405 3 - 4080.6 85 59 15 1591.2 57 79 16 1521.8 49 73 14 1368.0 Feed 290 3 - 2930.6 157 59 12 2281.2 46 105 17 1681.8 44 101 15 160* note: * samples spoiled prior to suiphite and thiosuiphate testing245Table 75: Sulphur Balance for Acid Phase Reactor pH Control Experimentsat 55 °CAcid Phase HRT Sulphur Compound 1mg S/i]Reactor p11 [days] SO42 S2 H2S g SO3 S203 TotalRecovered7.0 Feed 415 2 - 4 3 4240.6 224 113 31 8 12 388 921.2 216 112 38 8 15 389 921.8 241 116 30 13 34 434 1027.5 Feed 470 2 - 4 1 4770.6 282 61 13 2 13 371 781.2 270 61 16 2 15 364 761.8 235 83 23 2 23 366 778.0 Feed 396 3 - 4 1 4040.6 305 85 20 4 5 419 1041.2 242 109 21 6 26 404 1001.8 194 131 28 4 24 381 94246Table 76: Sulphur Balance for Molybdate Addition Experiments at 35 °CMolybdate HRT Sulphur Compound (mg S/I]Addition [days] S042 S2 H2S g S03 S203 Total %[mM] * * Recovered0.0 Feed 511 1- 512batch 1 0.6 209 67 18 2941.2 51 95 21 1671.8 56 97 22 1750.1 Feed 509 1- 510batch 1 0.6 66 104 22 1921.2 66 120 22 2081.8 61 127 24 2120.5 Feed 522 1 - 523batch 1 0.6 225 69 17 3111.2 47 111 21 1791.8 37 110 19 1660.75 Feed 470 3 - 510batch 1 0.6 315 61 15 1921.2 89 105 18 2081.8 85 113 21 2121.0 Feed 509 1 - 510batch 1 0.6 377 11 3 3911.2 305 30 6 3411.8 306 27 6 3391.0 Feed 416 3 - 419batch 2 0.6 379 14 4 3971.2 104 79 15 1981.8 92 80 16 1881.0 Feed 454 2 - 456batch3 0.6 386 39 8 4331.2 350 85 15 4501.8 336 102 19 457* note: * samples spoiled prior to sulphite and thiosuiphate testing247Table 77: Sulphur Balance for Iron and Molybdate Addition Experiments at 35 °CChemical HRT Sulphur Compound jmg S/I]Addition (days] SO42 2 H2S g S03 S203 Total %(mM] RecoveredOmMMo+ Feed 456 3 - 9 19 4875mg/I 0.6 219 186 32 2 20 459 94FeC136H2O 1.2 59 264 46 11 29 409 84(control) 1.8 68 266 49 8 25 416 85OmMMo Feed 454 2 - * * 456+5mg/I 0.6 386 39 8 * * 433FeCI36H2O 1.2 350 85 15 * * 4501.8 336 102 19 * * 4571.OOmMMo Feed 371 3 - 18 19 411+100mg/i 0.6 330 47 10 6 24 417 101FeC136H2O 1.2 249 89 16 13 46 413 1001.8 226 109 20 9 40 404 981.0 mM Mo Feed 446 3 - 11 14 474+200mg/i 0.6 248 139 28 17 47 479 101FeC136HO 1.2 58 213 35 18 45 369 781.8 0 232 43 9 50 334 70* note:’ * samples spoiled prior to suiphite and thiosuiphate testing248Table 78: Sulphur Balance for Molybdate Addition Experiments at 55 °CMolybdate HRT Sulphur Compound [mg S/I]Addition Fdaysl S042 52- H2S g S03 S203 Total %jmM] Recovered0 Feed 470 2 - 4 0 4760.6 282 61 18 2 13 376 791.2 268 62 16 3 15 364 761.8 235 83 23 2 9 352 740.5 Feed 493 3- 4 0 5000.6 326 60 15 2 10 413 831.2 316 63 16 3 18 416 831.8 311 63 16 2 6 398 801.0 Feed 457 3- 12 1 4730.6 320 49 14 4 13 400 841.2 306 47 11 6 25 395 841.8 306 55 15 4 5 385 81249Table 79: Sulphur Balance for Gas Stripping Experiments at 35 °CGas HRT Sulphur Compound [mg S/ilStripping [daysi S042 S2 H2S g S03 5203 Total %Experiment Recoveredwater Feed 467 3 - 11 14 495scrubber 0.6 127 186 38 17 47 415 84control 1.2 61 264 54 18 45 442 89batch 1 1.8 47 266 55 9 50 427 86ferric Feed 470 3 - 12 12 497chloride 0.6 376 69 56 2 30 533 107scrubber 1.2 169 108 21 3 25 326 66batch 1 1.8 94 107 21 3 12 237 48ferric Feed 190 3 - 8 2 203chloride 0.6 55 24 6 7 2 94 46scrubber 1.2 57 27 5 5 3 97 48batch2 1.8 57 32 6 5 5 105 52ferric Feed 390 3 - 4 1 398chloride 0.6 172 58 14 0 18 262 66scrubber 1.2 109 92 15 0 21 237 60batch 2 1.8 66 93 17 2 22 200 50+ so4250Table 80: Sulphur Balance for Gas Stripping Experiments at 55 °CGas HRT Sulphur Compound [mg S/IlStripping Idays] So42- 52 H2S g S03 S203 TotalExperiment Recoveredwater Feed 527 2 - 9 1 539scrubber 0.6 337 125 40 8 13 523 97control 1.2 327 124 33 8 25 517 961.8 411 71 26 6 5 519 96ferric Feed 513 3 - 6 0 522chloride 0.6 339 52 13 8 15 427 82scrubber 1.2 298 52 13 7 7 377 721.8 374 48 15 7 4 448 860251

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0058546/manifest

Comment

Related Items